TradingIQ - Counter Strike IQIntroducing "Counter Strike IQ" by TradingIQ
Counter Strike IQ is an exclusive trading algorithm developed by TradingIQ, designed to trade upside/downside breakouts of varying significance. By integrating artificial intelligence and IQ Technology, Counter Strike IQ analyzes historical and real-time price data to construct a dynamic trading system adaptable to various asset and timeframe combinations.
Philosophy of Counter Strike IQ
Counter Strike IQ operates on a single premise: Support and resistance levels cannot hold forever. At some point either side must break for the underlying asset to exhibit trends; otherwise, prices would be confined to an infinitely narrowing range.
Counter Strike IQ is designed to work straight out of the box. In fact, its simplicity requires just four user settings to manage output, making it incredibly straightforward to manage.
Minimum ATR Profit, Minimum ATR Stop, EMA Filter and EMA Filter Length are the only settings that manage the performance of Counter Strike IQ!
Traders don’t have to spend hours adjusting settings and trying to find what works best - Counter Strike IQ handles this on its own.
Key Features of Counter Strike IQ
Self-Learning Breakout Detection
Employs AI and IQ Technology to identify notable breakouts in real-time.
AI-Generated Trading Signals
Provides breakout trading signals derived from self-learning algorithms.
Comprehensive Trading System
Offers clear entry and exit labels.
Performance Tracking
Records and presents trading performance data, easily accessible for user analysis.
Self-Learning Trading Exits
Counter Strike IQ learns where to exit positions.
Long and Short Trading Capabilities
Supports both long and short positions to trade various market conditions.
Strike Channel
The Strike Channel represents what Counter Strike IQ considers a tradable long opportunity or a tradable short opportunity. The Strike Channel is dynamic and adjusts from chart to chart.
IQ Graph Gradient
Introduces the IQ Graph Gradient, designed to classify extreme values in price on a grand scale.
How It Works
Counter Strike IQ operates on a straightforward heuristic: go long during significant upside price moves that break established resistance levels and go short during significant downside price moves that break established support levels.
IQ Technology, TradingIQ's proprietary AI algorithm, defines what constitutes a “significant price move” and what’s considered a tradable breakout. For Counter Strike IQ, this algorithm evaluates all historical support/resistance breaks and any subsequent breakouts. For instance, the price move following up to a breakout is measured and learned from, including the significance of the identified support/resistance level (how long it’s been active, how far price moved away from it, etc). By analyzing these patterns, Counter Strike IQ adapts to identify and trade similar future breakout sequences.
In simple terms, Counter Strike IQ learns from violations of historical support/resistance levels to identify potential entry points at currently established support/resistance levels. Using this knowledge, it determines the optimal, current support/resistance price level where a breakout has a higher chance of occurring.
For long positions, Counter Strike IQ places a stop-market order at the AI-identified resistance point. If price violates this level a market order will be placed and a long position entered. Of course, this is how the algorithm trades, users can elect to use a stop-limit order amongst other order types for position entry. After the position is entered TP1 is placed (identifiable on the price chart). TP1 has a twofold purpose:
Acts as a legitimate profit target to exit 50% of the position.
Once TP1 is closed over, the initial stop loss is converted to a trailing stop, and the long position remains active so long as price continues to uptrend.
For short positions, Counter Strike IQ places a stop-market order at the AI-identified support point. If price violates this level a market order will be placed and a short position entered. Again, this is how the algorithm trades, users can elect to use a stop-limit order amongst other order types for position entry. Upon entry TP1 is placed (identifiable on the price chart). TP1 has a twofold purpose:
Acts as a legitimate profit target to exit 50% of the position.
Once TP1 is closed over, the initial stop loss is converted to a trailing stop, and the short position remains active so long as price continues to downtrend.
As a trading system, Counter Strike IQ exits TP1 using a limit order, with all stop losses exited as stop market orders.
What Classifies As a Tradable Upside Breakout or Tradable Downside Breakout?
For Counter Strike IQ, tradable price breakouts are not manually set but are instead learned by the system. What qualifies as a significant upside or downside breakout in one market might not hold the same significance in another. Counter Strike IQ continuously analyzes historical and current support/resistance levels, how far price has extended from those levels, the raw-dollar price move leading up to a violation of those levels, their longevity, and more, to determine which future levels have a higher chance of breaking out when retested!
The image above illustrates the Strike Channel and explains the corresponding prices and levels
The green upper line represents the Long Breakout Point.
The pink lower line represents the Short Breakout Point.
Any price between the two deviation points is considered “Acceptable”.
The image above shows a long position being entered after the Upside Breakout Point was reached.
Green arrows indicate that the strategy entered a long position at the highlighted price level.
Blue arrows indicate that the strategy exited a position, whether at TP1, the initial stop loss, or at the trailing stop.
Blue lines indicate the TP1 level for the current trade. Red lines indicate the initial stop loss price.
If price closes above TP1, the initial stop loss will be replaced with a trailing stop. A blue line (similar to the blue line shown for TP1) will trail price and correspond to the trailing stop price of the trade.
The image above shows the trailing stop price, represented by a blue line, used for the long position!
You can also hover over the trade labels to get more information about the trade—such as the entry price and exit price.
The image above shows a short position being entered after the Downside Breakout Point was reached.
Red arrows indicate that the strategy entered a short position at the highlighted price level.
Blue arrows indicate that the strategy exited a position, whether at TP1, the initial stop loss, or at the trailing stop.
Blue lines indicate the TP1 level for the current trade. Red lines indicate the initial stop loss price.
If price closes below TP1, the initial stop loss will be replaced with a trailing stop. A blue line (similar to the blue line shown for TP1) will trail price and correspond to the trailing stop price of the trade.
The image above shows the trailing stop price, represented by a blue line, used for the short position!
You can also hover over the trade labels to get more information about the trade—such as the entry price and exit price.
IQ Gradient Graph
The IQ Gradient Graph provides a macro characterization of extreme prices.
The lower macro extremity of the IQ Gradient Graph is colored green, while the upper macro extremity is colored red.
Minimum Profit Target And Stop Loss
The Minimum ATR Profit Target and Minimum ATR Stop Loss setting control the minimum allowed profit target and stop loss distance. On most timeframes users won’t have to alter these settings; however, on very-low timeframes such as the 1-minute chart, users can increase these values so gross profits exceed commission.
After changing either setting, Counter Strike IQ will retrain on historical data - accounting for the newly defined minimum profit target or stop loss.
AI Direction
The AI Direction setting controls the trade direction Counter Strike IQ is allowed to take.
“Trade Longs” allows for long trades.
“Trade Shorts” allows for short trades.
EMA Filter
The EMA Filter setting controls whether the AI should implement an EMA trading filter. Simply, if the EMA Filter is active, long trades can only initiate if price is trading above the user-defined EMA. Conversely, short trades can only initiate if price is trading below the user-defined EMA.
The image above shows the EMA Filter in action!
Verifying Counter Strike IQ’s Effectiveness
Counter Strike IQ automatically tracks its performance and displays the profit factor for the long strategy and the short strategy it uses. This information can be found in the table located in the top-right corner of your chart showing.
This table shows the long strategy profit factor and the short strategy profit factor.
The image above shows the long strategy profit factor and the short strategy profit factor for Counter Strike IQ.
A profit factor greater than 1 indicates a strategy profitably traded historical price data.
A profit factor less than 1 indicates a strategy unprofitably traded historical price data.
A profit factor equal to 1 indicates a strategy did not lose or gain money when trading historical price data.
Using Counter Strike IQ
While Counter Strike IQ is a full-fledged trading system with entries and exits - manual traders can certainly make use of its on chart indications and visualizations.
The hallmark feature of Counter Strike IQ is its ability to signal a breakout near its origin point. Long entries are often signaled near the start of a large upside price move; short entries are often signaled near the start of a large downside price move.
For live analysis, the Strike Channel serves as a valuable tool for identifying breakout points.
The further price moves toward the Upside Breakout Point (green), the stronger the indication that price might breakout to the upside. Conversely, the deeper price reaches toward the Downside Breakout Point (red), the stronger the indication that price might breakout to the downside.
Of course, should buying or selling pressure stall, price may fail to breakout at the identified breakout level. This is a natural consequence of any breakout trading strategy!
With this information at hand, traders can quickly switch between charts and timeframes to identify optimized areas of interest.
AI
Machine Learning Signal FilterIntroducing the "Machine Learning Signal Filter," an innovative trading indicator designed to leverage the power of machine learning to enhance trading strategies. This tool combines advanced data processing capabilities with user-friendly customization options, offering traders a sophisticated yet accessible means to optimize their market analysis and decision-making processes. Importantly, this indicator does not repaint, ensuring that signals remain consistent and reliable after they are generated.
Machine Learning Integration
The "Machine Learning Signal Filter" employs machine learning algorithms to analyze historical price data and identify patterns that may not be immediately apparent through traditional technical analysis. By utilizing techniques such as regression analysis and neural networks, the indicator continuously learns from new data, refining its predictive capabilities over time. This dynamic adaptability allows the indicator to adjust to changing market conditions, potentially improving the accuracy of trading signals.
Key Features and Benefits
Dynamic Signal Generation: The indicator uses machine learning to generate buy and sell signals based on complex data patterns. This approach enables it to adapt to evolving market trends, offering traders timely and relevant insights. Crucially, the indicator does not repaint, providing reliable signals that traders can trust.
Customizable Parameters: Users can fine-tune the indicator to suit their specific trading styles by adjusting settings such as the temporal synchronization and neural pulse rate. This flexibility ensures that the indicator can be tailored to different market environments.
Visual Clarity and Usability: The indicator provides clear visual cues on the chart, including color-coded signals and optional display of signal curves. Users can also customize the table's position and text size, enhancing readability and ease of use.
Comprehensive Performance Metrics: The indicator includes a detailed metrics table that displays key performance indicators such as return rates, trade counts, and win/loss ratios. This feature helps traders assess the effectiveness of their strategies and make data-driven decisions.
How It Works
The core of the "Machine Learning Signal Filter" is its ability to process and learn from large datasets. By applying machine learning models, the indicator identifies potential trading opportunities based on historical data patterns. It uses regression techniques to predict future price movements and neural networks to enhance pattern recognition. As new data is introduced, the indicator refines its algorithms, improving its accuracy and reliability over time.
Use Cases
Trend Following: Ideal for traders seeking to capitalize on market trends, the indicator helps identify the direction and strength of price movements.
Scalping: With its ability to provide quick signals, the indicator is suitable for scalpers aiming for rapid profits in volatile markets.
Risk Management: By offering insights into trade performance, the indicator aids in managing risk and optimizing trade setups.
In summary, the "Machine Learning Signal Filter" is a powerful tool that combines the analytical strength of machine learning with the practical needs of traders. Its ability to adapt and provide actionable insights makes it an invaluable asset for navigating the complexities of financial markets.
The "Machine Learning Signal Filter" is a tool designed to assist traders by providing insights based on historical data and machine learning techniques. It does not guarantee profitable trades and should be used as part of a comprehensive trading strategy. Users are encouraged to conduct their own research and consider their financial situation before making trading decisions. Trading involves significant risk, and it is possible to lose more than the initial investment. Always trade responsibly and be aware of the risks involved.
Support/Resistance v2 (ML) KmeanKmean with Standard Deviation Channel
1. Description of Kmean
Kmean (or K-means) is a popular clustering algorithm used to divide data into K groups based on their similarity. In the context of financial markets, Kmean can be applied to find the average price values over a specific period, allowing the identification of major trends and levels of support and resistance.
2. Application in Trading
In trading, Kmean is used to smooth out the price series and determine long-term trends. This helps traders make more informed decisions by avoiding noise and short-term fluctuations. Kmean can serve as a baseline around which other analytical tools, such as channels and bands, are constructed.
3. Description of Standard Deviation (stdev)
Standard deviation (stdev) is a statistical measure that indicates how much the values of data deviate from their mean value. In finance, standard deviation is often used to assess price volatility. A high standard deviation indicates strong price fluctuations, while a low standard deviation indicates stable movements.
4. Combining Kmean and Standard Deviation to Predict Short-Term Price Behavior
Combining Kmean and standard deviation creates a powerful tool for analyzing market conditions. Kmean shows the average price trend, while the standard deviation channels demonstrate the boundaries within which the price can fluctuate. This combination helps traders to:
Identify support and resistance levels.
Predict potential price reversals.
Assess risks and set stop-losses and take-profits.
Should you have any questions about code, please reach me at Tradingview directly.
Hope you find this script helpful!
Edge AI Forecast [Edge Terminal]This indicator inputs the previous 150 closing prices in a simple two-layer neural network, normalizes the network inputs using a sigmoid function, uses a feedforward calculation to send it to the second layer, shows the MSE loss curve and uses both automatic and manual backpropagation (user input) to find the most likely forecast values and uses the analog forecasting algorithm to adjust and optimize the data furthermore to display potential prices on the chart.
Here's how it works:
The idea behind this script is to train a simple neural network to predict the future x values based on the sample data. For this, we use 2 types of data, Price and Volume.
The thinking behind this is that price alone can’t be used in this case because it doesn’t provide enough meaningful pattern data for the network but price and volume together can change the game. We’re planning to use more different data sets and expand on this in the future.
To avoid a bad mix of results, we technically have two neural networks, each processing a different data type, one for volume data and one for price data.
The actual prediction is decided by the way price and volume of the closing price relate to each other. Basically, the network passes the price and volume and finds the best relation between the two data set outputs and predicts where the price could be based on the upcoming volume of the latest candle.
The network adjusts the weights and biases using optimization algorithms like gradient descent to minimize the difference between the predicted and actual stock prices, typically measured by a loss function, (in this case, mean squared error) which you can see using the error rate bubble.
This is a good measure to see how well the network is performing and the idea is to adjust the settings inputs such as learning rate, epochs and data source to get the lowest possible error rate. That’s when you’re getting the most accurate prediction results.
For each data set, we use a multi-layer network. In a multi-layer neural network, the outputs of neurons in one layer serve as inputs to neurons in the next layer. Initially, the input layer of the neural network receives the historical data. Each input neuron represents a feature, such as previous stock prices and trading volumes over a specific period.
The hidden layers perform feature extraction and transformation through a series of weighted connections and activation functions. Each neuron in a hidden layer computes a weighted sum of the inputs from the previous layer, applies an activation function to the sum, and passes the result to the next layer using the feedforward (activation) function.
For extraction, we use a normalization function. This function takes a value or data (such as bar price) and divides it up by max scale which is the highest possible value of the bar. The idea is to take a normalized number, which is either below 1 or under 2 for simple use in the neural network layers.
For the activation, after computing the weighted sum, the neuron applies an activation function a(x). To introduce non-linearity into the model to pass it to the next layer. We use sigmoid activation functions in this case. The main reason we use sigmoid function is because the resulting number is between 0 to 1 and is better for models where we have to predict the probability as an output.
The final output of the network is passed as an input to the analog forecasting function. This is an algorithm commonly used in weather prediction systems. In this case, this is used to make predictions by comparing current values and assuming the patterns might repeat in the future.
There are many different ways to build an analog forecasting function but in our case, we’re used similarity measurement model:
X, as the current situation or set of current variables.
Y, as the outcome or variable of interest.
Si as the historical situations or patterns, where i ranges from 1 to n.
Vi as the vector of variables describing historical situation Si.
Oi as the outcome associated with historical situation Si.
First, we define a similarity measure sim(X,Vi) that quantifies the similarity between the current situation X and historical situation Si based on their respective variables Vi.
Then we select the K most similar historical situations (KNN Machine learning) based on the similarity measure sim(X,Vi). We denote the rest of the selected historical situations as {Si1, Si2,...Sik).
Then we examine the outcomes associated with the selected historical situations {Oi1, Oi2,...,Oik}.
Then we use the outcomes of the selected historical situations to forecast the future outcome Y^ using weighted averaging.
Finally, the output value of the analog forecasting is standardized using a standardization function which is the opposite of the normalization function. This function takes a normalized number and turns it back to its original value by multiplying it by the max scale (highest value of the bar). This function is used when the final number is produced by the network output at the end of the analog forecasting to turn the final value back into a price so it can be displayed on the chart with PineScript.
Settings:
Data source: Source of the neural network's input data.
Sample Bars: How many historical bars do you want to input into the neural network
Prediction Bars: How many bars you want the script to forecast
Show Training Rate: This shows the neural network's error rate for the optimization phase
Learning Rate: how many times you want the script to change the model in response to the estimated error (automatic)
Epochs: the network cycle or how many times you want to run the data through the network from the first layer to the last one.
Usage:
The sample bars input determines the number of historical bars to be used as a reference for the network. You need to change the Epochs and Learning Rate inputs for each asset and chart timeframe to get the lowest error rate.
On the surface, the highest possible epoch and learning rate should produce the most effective results but that's not always the case.
If the epochs rate is too high, there is a chance we face overfitting. Essentially, you might be over processing good data which can make it useless.
On the other hand, if the learning rate is too high, the network may overshoot the optimal solution and diverge. This is almost like the same issue I mentioned above with a high epoch rate.
Access:
It took over 4 months to develop this script and we’re constantly improving it so it took a lot of manpower to develop this script. Also when it comes to neural networks, Pine Script isn’t the most optimal language to build a neural network in, so we had to resort to a few proprietary mathematical formulas to ensure this runs smoothly without giving out an error for overprocessing, specially when you have multiple neural networks with many layers.
The optimization done to make this script run on Pine Script is basically state of the art and because of this, we would like to keep the code closed source at the moment.
On the other hand we don’t want to publish the code publicly as we want to keep the trading edge this script gives us in a closed loop, for our own small group of members so we have to keep the code closed. We only accept invites from expert traders who understand how this script and algo trading works and the type of edge it provides.
Additionally, at the moment we don’t want to share the code as some of the parts of this network, specifically the way we hand the data from neural network output into the analog method formula are proprietary code and we’d like to keep it that way.
You can contact us for access and if we believe this works for your trading case, we will provide you with access.
MBAND 200 4H BTC/USDT - By MGS-TradingMBAND 200 4H BTC/USDT with RSI and Volume by MGS-Trading: A Neural Network-Inspired Indicator
Introduction:
The MBAND 200 4H BTC/USDT with RSI and Volume represents a groundbreaking achievement in the integration of artificial intelligence (AI) into cryptocurrency market analysis. Developed by MGS-Trading, this indicator is the culmination of extensive research and development efforts aimed at leveraging AI's power to enhance trading strategies. By synthesizing neural network concepts with traditional technical analysis, the MBAND indicator offers a dynamic, multi-dimensional view of the market, providing traders with unparalleled insights and actionable signals.
Innovative Approach:
Our journey to create the MBAND indicator began with a simple question: How can we mimic the decision-making prowess of a neural network in a trading indicator? The answer lay in the weighted aggregation of Exponential Moving Averages (EMAs) from multiple timeframes, each serving as a unique input akin to a neuron in a neural network. These weights are not arbitrary; they were painstakingly optimized through backtesting across various market conditions to ensure they reflect the significance of each timeframe’s contribution to overall market dynamics.
Core Features:
Neural Network-Inspired Weights: The heart of the MBAND indicator lies in its AI-inspired weighting system, which treats each timeframe’s EMA as an input node in a neural network. This allows the indicator to process complex market data in a nuanced and sophisticated manner, leading to more refined and informed trading signals.
Multi-Timeframe EMA Analysis: By analyzing EMAs from 15 minutes to 3 days, the MBAND indicator captures a comprehensive snapshot of market trends, enabling traders to make informed decisions based on a broad spectrum of data.
RSI and Volume Integration: The inclusion of the Relative Strength Index (RSI) and volume data adds layers of confirmation to the signals generated by the EMA bands. This multi-indicator approach helps in identifying high-probability setups, reinforcing the neural network’s concept of leveraging multiple data points for decision-making.
Usage Guidelines:
Signal Interpretation: The MBAND bands provide a visual representation of the market’s momentum and direction. A price moving above the upper band signals strength and potential continuation of an uptrend, while a move below the lower band suggests weakness and a possible downtrend.
Overbought/Oversold Conditions: The RSI component identifies when the asset is potentially overbought (>70) or oversold (<30). Traders should watch for these conditions near the MBAND levels for potential reversal opportunities.
Volume Confirmation: An increase in volume accompanying a price move towards or beyond an MBAND level serves as confirmation of the strength behind the move. This can indicate whether a breakout is likely to sustain or if a reversal has substantial backing.
Strategic Entry and Exit Points: Combine the MBAND readings with RSI and volume indicators to pinpoint strategic entry and exit points. For example, consider entering a long position when the price is near the lower MBAND, RSI indicates oversold conditions, and there is a notable volume increase.
About MGS-Trading:
At MGS-Trading, we are passionate about harnessing the transformative power of AI to revolutionize cryptocurrency trading. Our indicators and tools are designed to provide traders with advanced analytics and insights, drawing on the latest AI techniques and methodologies. The MBAND 200 4H BTC/USDT with RSI and Volume indicator is a prime example of our commitment to innovation, offering traders a sophisticated, AI-enhanced tool for navigating the complexities of the cryptocurrency markets.
Disclaimer:
The MBAND indicator is provided for informational purposes only and does not constitute investment advice. Trading cryptocurrencies involves significant risk and can result in the loss of your investment. We recommend conducting your own research and consulting with a qualified financial advisor before making any trading decisions.
AI Adaptive Money Flow Index (Clustering) [AlgoAlpha]🌟🚀 Dive into the future of trading with our latest innovation: the AI Adaptive Money Flow Index by AlgoAlpha Indicator! 🚀🌟
Developed with the cutting-edge power of Machine Learning, this indicator is designed to revolutionize the way you view market dynamics. 🤖💹 With its unique blend of traditional Money Flow Index (MFI) analysis and advanced k-means clustering, it adapts to market conditions like never before.
Key Features:
📊 Adaptive MFI Analysis: Utilizes the classic MFI formula with a twist, adjusting its parameters based on AI-driven clustering.
🧠 AI-Driven Clustering: Applies k-means clustering to identify and adapt to market states, optimizing the MFI for current conditions.
🎨 Customizable Appearance: Offers adjustable settings for overbought, neutral, and oversold levels, as well as colors for uptrends and downtrends.
🔔 Alerts for Key Market Movements: Set alerts for trend reversals, overbought, and oversold conditions, ensuring you never miss a trading opportunity.
Quick Guide to Using the AI Adaptive MFI (Clustering):
🛠 Customize the Indicator: Customize settings like MFI source, length, and k-means clustering parameters to suit your analysis.
📈 Market Analysis: Monitor the dynamically adjusted overbought, neutral, and oversold levels for insights into market conditions. Watch for classification symbols ("+", "0", "-") for immediate understanding of the current market state. Look out for reversal signals (▲, ▼) to get potential entry points.
🔔 Set Alerts: Utilize the built-in alert conditions for trend changes, overbought, and oversold signals to stay ahead, even when you're not actively monitoring the charts.
How It Works:
The AI Adaptive Money Flow Index employs the k-means clustering machine learning algorithm to refine the traditional Money Flow Index, dynamically adjusting overbought, neutral, and oversold levels based on market conditions. This method analyzes historical MFI values, grouping them into initial clusters using the traditional MFI's overbought, oversold and neutral levels, and then finding the mean of each cluster, which represent the new market states thresholds. This adaptive approach ensures the indicator's sensitivity in real-time, offering a nuanced understanding of market trend and volume analysis.
By recalibrating MFI thresholds for each new data bar, the AI Adaptive MFI intelligently conforms to changing market dynamics. This process, assessing past periods to adjust the indicator's parameters, provides traders with insights finely tuned to recent market behavior. Such innovation enhances decision-making, leveraging the latest data to inform trading strategies. 🌐💥
Machine Learning: Multiple Logistic Regression
Multiple Logistic Regression Indicator
The Logistic Regression Indicator for TradingView is a versatile tool that employs multiple logistic regression based on various technical indicators to generate potential buy and sell signals. By utilizing key indicators such as RSI, CCI, DMI, Aroon, EMA, and SuperTrend, the indicator aims to provide a systematic approach to decision-making in financial markets.
How It Works:
Technical Indicators:
The script uses multiple technical indicators such as RSI, CCI, DMI, Aroon, EMA, and SuperTrend as input variables for the logistic regression model.
These indicators are normalized to create categorical variables, providing a consistent scale for the model.
Logistic Regression:
The logistic regression function is applied to the normalized input variables (x1 to x6) with user-defined coefficients (b0 to b6).
The logistic regression model predicts the probability of a binary outcome, with values closer to 1 indicating a bullish signal and values closer to 0 indicating a bearish signal.
Loss Function (Cross-Entropy Loss):
The cross-entropy loss function is calculated to quantify the difference between the predicted probability and the actual outcome.
The goal is to minimize this loss, which essentially measures the model's accuracy.
// Error Function (cross-entropy loss)
loss(y, p) =>
-y * math.log(p) - (1 - y) * math.log(1 - p)
// y - depended variable
// p - multiple logistic regression
Gradient Descent:
Gradient descent is an optimization algorithm used to minimize the loss function by adjusting the weights of the logistic regression model.
The script iteratively updates the weights (b1 to b6) based on the negative gradient of the loss function with respect to each weight.
// Adjusting model weights using gradient descent
b1 -= lr * (p + loss) * x1
b2 -= lr * (p + loss) * x2
b3 -= lr * (p + loss) * x3
b4 -= lr * (p + loss) * x4
b5 -= lr * (p + loss) * x5
b6 -= lr * (p + loss) * x6
// lr - learning rate or step of learning
// p - multiple logistic regression
// x_n - variables
Learning Rate:
The learning rate (lr) determines the step size in the weight adjustment process. It prevents the algorithm from overshooting the minimum of the loss function.
Users can set the learning rate to control the speed and stability of the optimization process.
Visualization:
The script visualizes the output of the logistic regression model by coloring the SMA.
Arrows are plotted at crossover and crossunder points, indicating potential buy and sell signals.
Lables are showing logistic regression values from 1 to 0 above and below bars
Table Display:
A table is displayed on the chart, providing real-time information about the input variables, their values, and the learned coefficients.
This allows traders to monitor the model's interpretation of the technical indicators and observe how the coefficients change over time.
How to Use:
Parameter Adjustment:
Users can adjust the length of technical indicators (rsi_length, cci_length, etc.) and the Z score length based on their preference and market characteristics.
Set the initial values for the regression coefficients (b0 to b6) and the learning rate (lr) according to your trading strategy.
Signal Interpretation:
Buy signals are indicated by an upward arrow (▲), and sell signals are indicated by a downward arrow (▼).
The color-coded SMA provides a visual representation of the logistic regression output by color.
Table Information:
Monitor the table for real-time information on the input variables, their values, and the learned coefficients.
Keep an eye on the learning rate to ensure a balance between model adjustment speed and stability.
Backtesting and Validation:
Before using the script in live trading, conduct thorough backtesting to evaluate its performance under different market conditions.
Validate the model against historical data to ensure its reliability.
Machine Learning: STDEV Oscillator [YinYangAlgorithms]This Indicator aims to fill a gap within traditional Standard Deviation Analysis. Rather than its usual applications, this Indicator focuses on applying Standard Deviation within an Oscillator and likewise applying a Machine Learning approach to it. By doing so, we may hope to achieve an Adaptive Oscillator which can help display when the price is deviating from its standard movement. This Indicator may help display both when the price is Overbought or Underbought, and likewise, where the price may face Support and Resistance. The reason for this is that rather than simply plotting a Machine Learning Standard Deviation (STDEV), we instead create a High and a Low variant of STDEV, and then use its Highest and Lowest values calculated within another Deviation to create Deviation Zones. These zones may help to display these Support and Resistance locations; and likewise may help to show if the price is Overbought or Oversold based on its placement within these zones. This Oscillator may also help display Momentum when the High and/or Low STDEV crosses the midline (0). Lastly, this Oscillator may also be useful for seeing the spacing between the High and Low of the STDEV; large spacing may represent volatility within the STDEV which may be helpful for seeing when there is Momentum in the form of volatility.
Tutorial:
Above is an example of how this Indicator looks on BTC/USDT 1 Day. As you may see, when the price has parabolic movement, so does the STDEV. This is due to this price movement deviating from the mean of the data. Therefore when these parabolic movements occur, we create the Deviation Zones accordingly, in hopes that it may help to project future Support and Resistance locations as well as helping to display when the price is Overbought and Oversold.
If we zoom in a little bit, you may notice that the Support Zone (Blue) is smaller than the Resistance Zone (Orange). This is simply because during the last Bull Market there was more parabolic price deviation than there was during the Bear Market. You may see this if you refer to their values; the Resistance Zone goes to ~18k whereas the Support Zone is ~10.5k. This is completely normal and the way it is supposed to work. Due to the nature of how STDEV works, this Oscillator doesn’t use a 1:1 ratio and instead can develop and expand as exponential price action occurs.
The Neutral (0) line may also act as a Support and Resistance location. In the example above we can see how when the STDEV is below it, it acts as Resistance; and when it’s above it, it acts as Support.
This Neutral line may also provide us with insight as towards the momentum within the market and when it has shifted. When the STDEV is below the Neutral line, the market may be considered Bearish. When the STDEV is above the Neutral line, the market may be considered Bullish.
The Red Line represents the STDEV’s High and the Green Line represents the STDEV’s Low. When the STDEV’s High and Low get tight and close together, this may represent there is currently Low Volatility in the market. Low Volatility may cause consolidation to occur, however it also leaves room for expansion.
However, when the STDEV’s High and Low are quite spaced apart, this may represent High levels of Volatility in the market. This may mean the market is more prone to parabolic movements and expansion.
We will conclude our Tutorial here. Hopefully this has given you some insight into how applying Machine Learning to a High and Low STDEV then creating Deviation Zones based on it may help project when the Momentum of the Market is Bullish or Bearish; likewise when the price is Overbought or Oversold; and lastly where the price may face Support and Resistance in the form of STDEV.
If you have any questions, comments, ideas or concerns please don't hesitate to contact us.
HAPPY TRADING!
Machine Learning: Optimal Length [YinYangAlgorithms]This Indicator aims to solve an issue that most others face; static lengths. This Indicator will scan lengths from the Min to Max setting (1 - 400 by default) to calculate which is the most Optimal Length in the current market condition. Almost every Indicator uses a length in some part of their calculation, and this length is usually adjustable via the Settings; however it is generally a static fixed length. Static non changing lengths may not always produce optimal results. As market conditions change generally the optimal length will too. For this reason we have created this indicator.
This Indicator will create a Neutral (Min - Max Length), Fast (Min - Mid Length ((Max - Min) / 2)) and Slow (Mid Length ((Max - Min) / 2) - Max Length). This allows you to understand which the Optimal Fast, Slow and Neutral lengths are within the given Mix and Max length settings.
This Indicator then plots these Optimal Lengths as an Oscillator which can then be used within ANOTHER Indicator as a Source within its Settings. Stand alone this Indicator may not prove all that useful, however when its Lengths are inputted into another Indicator it may prove very useful. This allows other Indicators to use the Optimal Length within its calculations from the Settings rather than relying on simply a fixed length. Unfortunately this results in users needing to manually plug the Optimal Length plots into the second Indicator; but it also allows for endless possibilities with applying Machine Learning Optimal Lengths within both Traditional and Non-Traditional Indicators and may give other Pine Coders an easy and effective way to add Machine Learning auto adjustable lengths within their already created Indicators.
The beautiful part about this Indicator is that aside from inputting the Optimal Length Plot into another Indicator, there is no manual updating needed. When the Optimal Length changes, the change will automatically reflect in the other Indicator without the need for you to manually adjust its length. This may be very useful with both time preservation, as well as if there is an automated strategy based upon said Indicator that now won’t need manual intervention.
Tutorial:
By default this is what the Machine Learning: Optimal Length Indicator looks like. It is simply a way of both Displaying and Plotting our current Optimal Length so that we may then use it as a source within ANOTHER Indicator. This will allow the automation of an Optimal Length to be updated, rather than needing any manual input from yourself (aside from set up).
For instance if you set the start length to 1 and the end length to 400 (default settings), it will scan to find the optimal Length setting between 1 and 400. This features 3 types of lengths:
Fast (Green Line): 1-199 (from start length to half way of total)
Slow (Red Line): 200 - 400 (mid way to end length)
Neutral (Blue Line): 1 - 400 (start to end length)
By breaking down the Optimal Length detection into these 3 different types, we can see how the Optimal Length compares and changes based on the lengths allotted to them and how performance changes.
For instance, you may notice that both the Fast and Slow Optimal Length didn’t change much in the example above; however the Neutral Optimal Length changed quite a bit. This is due to the fact that the Neutral is inclusive of all lengths available and may be considered the more accurate due to that. However, this doesn’t mean the Fast and Slow lengths aren’t important and should be used. They may be useful for seeing how something fairs in a Fast and Slow standpoint.
If you change your TimeFrame from 15 minute to 1 Day, you’ll notice that the Optimal Lengths gravitate towards their upper bounds:
199 is max for Fast, it’s at 195
400 is max for Slow, its at 393
400 is max for Neutral, its at 399
The Optimal Length may move up to its upper bounds on Higher Time Frames because there is a lot of price action and long term data being displayed. This may lead to higher lengths performing better in a profitability standpoint since its data is based on so far back and such drastic price movements.
Below we’re going to go through a few examples, including the code so you may reproduce the example and have an understanding of how versatile Inputting an Optimal Length as a source may be within Traditional Indicators.
Adding the Machine Learning: Optimal Length to another Indicator:
You may add the Optimal Length to another Indicator as shown in the example above. In the example we are adding the ‘Machine Learning: Optimal Length - Neutral’ to our Neutral Length within the Settings. The external Indicator needs to have the ability to input the Optimal Length as a Source, this way it can automatically change within the external Indicator when the Optimal Length Indicator changes its Optimal Length.
Please note you may get an error within an external Indicator that accepts the Length as a Source if you don’t select the Machine Learning: Optimal Length. For instance, if you use ‘Close’ within BTC/USDT the length used would be ~36,000. This length is too long and will throw an error.
For this reason, we will ensure the Max Length that may be used is 1000.
Please note, on lower Time Frames you may need to adjust the Max Length. For instance if 20k bar data is used, the Max Length ‘may’ fail to load when going by default Min: 1 and Max: 400. Generally with most pairs it will load if your TradingView subscription is Premium or greater; however if it is less there is a chance it may fail. If it fails for you too often please lower the Max Length Amount; or send us a message we can look into a fix for this.
*** If it fails to load, please try removing the external Indicator and re-adding it and adding the Lengths back as a Source within the Settings. Sometimes it fails, but re-adding may fix it. If it keeps failing afterwards, reduce the Max Length Amount as mentioned above. ***
Simple Moving Average:
In this example above have the Fast, Slow and Neutral Optimal Length formatted as a Slow Moving Average. The first example is on the 15 minute Time Frame and the second is on the 1 Day Time Frame, demonstrating how the length changes based on the Time Frame and the effects it may have.
Here is the code for the example Indicator shown above. This example shows how you may use the Optimal Length as a Source and then use that Optimal Length and plot it as a Simple Moving Average:
//@version=5
indicator("Optimal Length - Backtesting - MA", overlay=true, max_bars_back=5000)
outputType = input.string("All", "Output Type", options= )
lengthSource = input.source(close, "Neutral Length")
lengthSource_fast = input.source(close, "Fast Length")
lengthSource_slow = input.source(close, "Slow Length")
showNeutral = outputType == "Neutral" or outputType == "Fast + Neutral" or outputType == "Slow + Neutral" or outputType == "All"
showFast = outputType == "Fast" or outputType == "Fast + Neutral" or outputType == "Fast + Slow" or outputType == "All"
showSlow = outputType == "Slow" or outputType == "Slow + Neutral" or outputType == "Fast + Slow" or outputType == "All"
//Neutral
optimalLength = math.min(math.max(math.round(lengthSource), 1), 1000)
optimalMA = ta.sma(close, optimalLength)
//Fast
optimalLength_fast = math.min(math.max(math.round(lengthSource_fast), 1), 1000)
optimalMA_fast = ta.sma(close, optimalLength_fast)
//Slow
optimalLength_slow = math.min(math.max(math.round(lengthSource_slow), 1), 1000)
optimalMA_slow = ta.sma(close, optimalLength_slow)
plot(showNeutral ? optimalMA : na, color=color.blue)
plot(showFast ? optimalMA_fast : na, color=color.green)
plot(showSlow ? optimalMA_slow : na, color=color.red)
Bollinger Bands:
In the two examples above for Bollinger Bands we have first the 15 Minute Time Frame and then the 1 Day Time Frame. As described above in ‘Adding the Machine Learning: Optimal Length to another Indicator’ sometimes it may fail to load, for this reason in the 15 Minute it was reduced to a max of 300 Length.
Bollinger Bands are a way to see a Simple Moving Average (SMA) that then uses Standard Deviation to identify how much deviation has occurred. This Deviation is than Added and Subtracted from the SMA to create the Bollinger Bands which help Identify possible movement zones that are ‘within range’. This may mean that the price may face Support / Resistance when it reaches the Outer / Inner bounds of the Bollinger Bands. Likewise, it may mean the Price is ‘Overbought’ when outside and above or ‘Underbought’ when outside and below the Bollinger Bands.
By applying All 3 different types of Optimal Lengths towards a Traditional Bollinger Band calculation we may hope to see different ranges of Bollinger Bands and how different lookback lengths may imply possible movement ranges on both a Short Term, Long Term and Neutral perspective. By seeing these possible ranges you may have the ability to identify more levels of Support and Resistance over different lengths and Trading Styles.
Below is the code for the Bollinger Bands example above:
//@version=5
indicator("Optimal Length - Backtesting - Bollinger Bands", overlay=true, max_bars_back=5000)
outputType = input.string("All", "Output Type", options= )
lengthSource = input.source(close, "Neutral Length")
lengthSource_fast = input.source(close, "Fast Length")
lengthSource_slow = input.source(close, "Slow Length")
showNeutral = outputType == "Neutral" or outputType == "Fast + Neutral" or outputType == "Slow + Neutral" or outputType == "All"
showFast = outputType == "Fast" or outputType == "Fast + Neutral" or outputType == "Fast + Slow" or outputType == "All"
showSlow = outputType == "Slow" or outputType == "Slow + Neutral" or outputType == "Fast + Slow" or outputType == "All"
mult = 2.0
src = close
neutralColor = color.blue
slowColor = color.red
fastColor = color.green
//Neutral
optimalLength = math.min(math.max(math.round(lengthSource), 1), 1000)
optimalMA = ta.sma(close, optimalLength)
//Fast
optimalLength_fast = math.min(math.max(math.round(lengthSource_fast), 1), 1000)
optimalMA_fast = ta.sma(close, optimalLength_fast)
//Slow
optimalLength_slow = math.min(math.max(math.round(lengthSource_slow), 1), 1000)
optimalMA_slow = ta.sma(close, optimalLength_slow)
//Neutral Bollinger Bands
dev = mult * ta.stdev(src, math.round(optimalLength))
upper = optimalMA + dev
lower = optimalMA - dev
plot(showNeutral ? optimalMA : na, "Neutral Basis", color=color.new(neutralColor, 0))
p1 = plot(showNeutral ? upper : na, "Neutral Upper", color=color.new(neutralColor, 50))
p2 = plot(showNeutral ? lower : na, "Neutral Lower", color=color.new(neutralColor, 50))
fill(p1, p2, title = "Neutral Background", color=color.new(neutralColor, 96))
//Slow Bollinger Bands
dev_slow = mult * ta.stdev(src, math.round(optimalLength_slow))
upper_slow = optimalMA_slow + dev_slow
lower_slow = optimalMA_slow - dev_slow
plot(showFast ? optimalMA_slow : na, "Slow Basis", color=color.new(slowColor, 0))
p1_slow = plot(showFast ? upper_slow : na, "Slow Upper", color=color.new(slowColor, 50))
p2_slow = plot(showFast ? lower_slow : na, "Slow Lower", color=color.new(slowColor, 50))
fill(p1_slow, p2_slow, title = "Slow Background", color=color.new(slowColor, 96))
//Fast Bollinger Bands
dev_fast = mult * ta.stdev(src, math.round(optimalLength_fast))
upper_fast = optimalMA_fast + dev_fast
lower_fast = optimalMA_fast - dev_fast
plot(showSlow ? optimalMA_fast : na, "Fast Basis", color=color.new(fastColor, 0))
p1_fast = plot(showSlow ? upper_fast : na, "Fast Upper", color=color.new(fastColor, 50))
p2_fast = plot(showSlow ? lower_fast : na, "Fast Lower", color=color.new(fastColor, 50))
fill(p1_fast, p2_fast, title = "Fast Background", color=color.new(fastColor, 96))
Donchian Channels:
Above you’ll see two examples of Machine Learning: Optimal Length applied to Donchian Channels. These are displayed with both the 15 Minute Time Frame and the 1 Day Time Frame.
Donchian Channels are a way of seeing potential Support and Resistance within a given lookback length. They are a way of withholding the High’s and Low’s of a specific lookback length and looking for deviation within this length. By applying our Fast, Slow and Neutral Machine Learning: Optimal Length to these Donchian Channels way may hope to achieve a viable range of High’s and Low’s that one may use to Identify Support and Resistance locations for different ranges of Optimal Lengths and likewise potentially different Trading Strategies.
The code to reproduce these Donchian Channels as displayed above is so:
//@version=5
indicator("Optimal Length - Backtesting - Donchian Channels", overlay=true, max_bars_back=5000)
outputType = input.string("All", "Output Type", options= )
lengthSource = input.source(close, "Neutral Length")
lengthSource_fast = input.source(close, "Fast Length")
lengthSource_slow = input.source(close, "Slow Length")
showNeutral = outputType == "Neutral" or outputType == "Fast + Neutral" or outputType == "Slow + Neutral" or outputType == "All"
showFast = outputType == "Fast" or outputType == "Fast + Neutral" or outputType == "Fast + Slow" or outputType == "All"
showSlow = outputType == "Slow" or outputType == "Slow + Neutral" or outputType == "Fast + Slow" or outputType == "All"
mult = 2.0
src = close
neutralColor = color.blue
slowColor = color.red
fastColor = color.green
//Neutral
optimalLength = math.min(math.max(math.round(lengthSource), 1), 1000)
optimalMA = ta.sma(close, optimalLength)
//Fast
optimalLength_fast = math.min(math.max(math.round(lengthSource_fast), 1), 1000)
optimalMA_fast = ta.sma(close, optimalLength_fast)
//Slow
optimalLength_slow = math.min(math.max(math.round(lengthSource_slow), 1), 1000)
optimalMA_slow = ta.sma(close, optimalLength_slow)
//Neutral Donchian Channels
lower_dc = ta.lowest(optimalLength)
upper_dc = ta.highest(optimalLength)
basis_dc = math.avg(upper_dc, lower_dc)
plot(showNeutral ? basis_dc : na, "Donchain Channel - Neutral Basis", color=color.new(neutralColor, 0))
u = plot(showNeutral ? upper_dc : na, "Donchain Channel - Neutral Upper", color=color.new(neutralColor, 50))
l = plot(showNeutral ? lower_dc : na, "Donchain Channel - Neutral Lower", color=color.new(neutralColor, 50))
fill(u, l, color=color.new(neutralColor, 96), title = "Donchain Channel - Neutral Background")
//Fast Donchian Channels
lower_dc_fast = ta.lowest(optimalLength_fast)
upper_dc_fast = ta.highest(optimalLength_fast)
basis_dc_fast = math.avg(upper_dc_fast, lower_dc_fast)
plot(showFast ? basis_dc_fast : na, "Donchain Channel - Fast Neutral Basis", color=color.new(fastColor, 0))
u_fast = plot(showFast ? upper_dc_fast : na, "Donchain Channel - Fast Upper", color=color.new(fastColor, 50))
l_fast = plot(showFast ? lower_dc_fast : na, "Donchain Channel - Fast Lower", color=color.new(fastColor, 50))
fill(u_fast, l_fast, color=color.new(fastColor, 96), title = "Donchain Channel - Fast Background")
//Slow Donchian Channels
lower_dc_slow = ta.lowest(optimalLength_slow)
upper_dc_slow = ta.highest(optimalLength_slow)
basis_dc_slow = math.avg(upper_dc_slow, lower_dc_slow)
plot(showSlow ? basis_dc_slow : na, "Donchain Channel - Slow Neutral Basis", color=color.new(slowColor, 0))
u_slow = plot(showSlow ? upper_dc_slow : na, "Donchain Channel - Slow Upper", color=color.new(slowColor, 50))
l_slow = plot(showSlow ? lower_dc_slow : na, "Donchain Channel - Slow Lower", color=color.new(slowColor, 50))
fill(u_slow, l_slow, color=color.new(slowColor, 96), title = "Donchain Channel - Slow Background")
Envelopes / Envelopes Adjusted:
Envelopes are an interesting one in the sense that they both may be perceived as useful; however we deem that with the use of an ‘Optimal Length’ that the ‘Envelopes Adjusted’ may work best. We will start with examples of the Traditional Envelope then showcase the Adjusted version.
Envelopes:
As you may see, a Traditional form of Envelopes even produced with our Machine Learning: Optimal Length may not produce optimal results. Unfortunately this may occur with some Traditional Indicators and they may need some adjustments as you’ll notice with the ‘Envelopes Adjusted’ version. However, even without the adjustments, these Envelopes may be useful for seeing ‘Overbought’ and ‘Oversold’ locations within a Machine Learning: Optimal Length standpoint.
Envelopes Adjusted:
By adding an adjustment to these Envelopes, we may hope to better reflect out Optimal Length within it. This is caused by adding a ratio reflection towards the current length of the Optimal Length and the max Length used. This allows for the Fast and Neutral (and potentially Slow if Neutral is greater) to achieve a potentially more accurate result.
Envelopes, much like Bollinger Bands are a way of seeing potential movement zones along with potential Support and Resistance. However, unlike Bollinger Bands which are based on Standard Deviation, Envelopes are based on percentages +/- from the Simple Moving Average.
The code used to reproduce the example above is as follows:
//@version=5
indicator("Optimal Length - Backtesting - Envelopes", overlay=true, max_bars_back=5000)
outputType = input.string("All", "Output Type", options= )
displayType = input.string("Envelope Adjusted", "Display Type", options= )
lengthSource = input.source(close, "Neutral Length")
lengthSource_fast = input.source(close, "Fast Length")
lengthSource_slow = input.source(close, "Slow Length")
showNeutral = outputType == "Neutral" or outputType == "Fast + Neutral" or outputType == "Slow + Neutral" or outputType == "All"
showFast = outputType == "Fast" or outputType == "Fast + Neutral" or outputType == "Fast + Slow" or outputType == "All"
showSlow = outputType == "Slow" or outputType == "Slow + Neutral" or outputType == "Fast + Slow" or outputType == "All"
mult = 2.0
src = close
neutralColor = color.blue
slowColor = color.red
fastColor = color.green
//Neutral
optimalLength = math.min(math.max(math.round(lengthSource), 1), 1000)
optimalMA = ta.sma(close, optimalLength)
//Fast
optimalLength_fast = math.min(math.max(math.round(lengthSource_fast), 1), 1000)
optimalMA_fast = ta.sma(close, optimalLength_fast)
//Slow
optimalLength_slow = math.min(math.max(math.round(lengthSource_slow), 1), 1000)
optimalMA_slow = ta.sma(close, optimalLength_slow)
percent = 10.0
maxAmount = math.max(optimalLength, optimalLength_fast, optimalLength_slow)
//Neutral
k = displayType == "Envelope" ? percent/100.0 : (percent/100.0) / (optimalLength / maxAmount)
upper_env = optimalMA * (1 + k)
lower_env = optimalMA * (1 - k)
plot(showNeutral ? optimalMA : na, "Envelope - Neutral Basis", color=color.new(neutralColor, 0))
u_env = plot(showNeutral ? upper_env : na, "Envelope - Neutral Upper", color=color.new(neutralColor, 50))
l_env = plot(showNeutral ? lower_env : na, "Envelope - Neutral Lower", color=color.new(neutralColor, 50))
fill(u_env, l_env, color=color.new(neutralColor, 96), title = "Envelope - Neutral Background")
//Fast
k_fast = displayType == "Envelope" ? percent/100.0 : (percent/100.0) / (optimalLength_fast / maxAmount)
upper_env_fast = optimalMA_fast * (1 + k_fast)
lower_env_fast = optimalMA_fast * (1 - k_fast)
plot(showFast ? optimalMA_fast : na, "Envelope - Fast Basis", color=color.new(fastColor, 0))
u_env_fast = plot(showFast ? upper_env_fast : na, "Envelope - Fast Upper", color=color.new(fastColor, 50))
l_env_fast = plot(showFast ? lower_env_fast : na, "Envelope - Fast Lower", color=color.new(fastColor, 50))
fill(u_env_fast, l_env_fast, color=color.new(fastColor, 96), title = "Envelope - Fast Background")
//Slow
k_slow = displayType == "Envelope" ? percent/100.0 : (percent/100.0) / (optimalLength_slow / maxAmount)
upper_env_slow = optimalMA_slow * (1 + k_slow)
lower_env_slow = optimalMA_slow * (1 - k_slow)
plot(showSlow ? optimalMA_slow : na, "Envelope - Slow Basis", color=color.new(slowColor, 0))
u_env_slow = plot(showSlow ? upper_env_slow : na, "Envelope - Slow Upper", color=color.new(slowColor, 50))
l_env_slow = plot(showSlow ? lower_env_slow : na, "Envelope - Slow Lower", color=color.new(slowColor, 50))
fill(u_env_slow, l_env_slow, color=color.new(slowColor, 96), title = "Envelope - Slow Background")
Hopefully these examples, including reproducing code, have given you some insight as to how useful this Machine Learning: Optimal Length may be and how another Indicator may easily modify their existing code to incorporate the usage of such Machine Learning: Optimal Length. We likewise will publish a Backtesting Indicator which incorporates all of the concepts we’ve gone over within here; in case you wish to take advantage of the Traditional Indicators mentioned above that allow the input of Machine Learning: Optimal Length and don’t wish to code them.
If you have any questions, comments, ideas or concerns please don't hesitate to contact us.
HAPPY TRADING!
Backtest Strategy Optimizer AdapterBacktest Strategy Optimizer Adapter
With this library, you will be able to run one or multiple backtests with different variables (combinations). For example, you can run 100 backtests of Supertrend at once with an increment factor of 0.1. This way, you can easily fetch the most profitable settings and apply them to your strategy.
To get a better understanding of the code, you can check the code below.
Single backtest results
= backtest.results(date_start, date_end, long_entry, long_exit, take_profit_percentage, stop_loss_percentage, atr_length, initial_capital, order_size, commission)
Add backtest results to a table
backtest.table(initial_capital, profit_and_loss, open_balance, winrate, entries, exits, wins, losses, backtest_table_position, backtest_table_margin, backtest_table_transparency, backtest_table_cell_color, backtest_table_title_cell_color, backtest_table_text_color)
Backtest result without chart labels
= backtest.run(date_start, date_end, long_entry, long_exit, take_profit_percentage, stop_loss_percentage, atr_length, initial_capital, order_size, commission)
Backtest result profit
profit = backtest.profit(date_start, date_end, long_entry, long_exit, take_profit_percentage, stop_loss_percentage, atr_length, initial_capital, order_size, commission)
Backtest result winrate
winrate = backtest.winrate(date_start, date_end, long_entry, long_exit, take_profit_percentage, stop_loss_percentage, atr_length, initial_capital, order_size, commission)
Start Date
You can set the start date either by using a timestamp or a number that refers to the number of bars back.
Stop Loss / Take Profit Issue
Unfortunately, I did not manage to achieve 100% accuracy for the take profit and stop loss. The original TradingView backtest can stop at the correct position within a bar using the strategy.exit stop and limit variables. However, it seems unachievable with a crossunder/crossover function in PineScript unless it is calculated on every tick (which would make the backtesting results invalid). So far, I have not found a workaround, and I would be grateful if someone could solve this issue, if it is even possible. If you have any solutions or fixes, please let me know!
Multiple Backtest Results / Optimizer
You can run multiple backtests in a single strategy or indicator, but there are certain requirements for placing the correct code in the right way. To view examples of running multiple backtests, you can refer to the links provided in the updates I posted below. In the samples I have also explained how you can auto-generate code for your backtest strategy.
Machine Learning: Anchored Gaussian Process Regression [LuxAlgo]Machine Learning: Anchored Gaussian Process Regression is an anchored version of Machine Learning: Gaussian Process Regression .
It implements Gaussian Process Regression (GPR), a popular machine-learning method capable of estimating underlying trends in prices as well as forecasting them. Users can set a Training Window by choosing 2 points. GPR will be calculated for the data between these 2 points.
Do remember that forecasting trends in the market is challenging, do not use this tool as a standalone for your trading decisions.
🔶 USAGE
When adding the indicator to the chart, users will be prompted to select a starting and ending point for the calculations, click on your chart to select those points.
Start & end point are named 'Anchor 1' & 'Anchor 2', the Training Window is located between these 2 points. Once both points are positioned, the Training Window is set, whereafter the Gaussian Process Regression (GPR) is calculated using data between both Anchors .
The blue line is the GPR fit, the red line is the GPR prediction, derived from data between the Training Window .
Two user settings controlling the trend estimate are available, Smooth and Sigma.
Smooth determines the smoothness of our estimate, with higher values returning smoother results suitable for longer-term trend estimates.
Sigma controls the amplitude of the forecast, with values closer to 0 returning results with a higher amplitude.
One of the advantages of the anchoring process is the ability for the user to evaluate the accuracy of forecasts and further understand how settings affect their accuracy.
The publication also shows the mean average (faint silver line), which indicates the average of the prices within the calculation window (between the anchors). This can be used as a reference point for the forecast, seeing how it deviates from the training window average.
🔶 DETAILS
🔹 Limited Training Window
The Training Window is limited due to matrix.new() limitations in size.
When the 2 points are too far from each other (as in the latter example), the line will end at the maximum limit, without giving a size error.
The red forecasted line is always given priority.
🔹 Positioning Anchors
Typically Anchor 1 is located further in history than Anchor 2 , however, placing Anchor 2 before Anchor 1 is perfectly possibly, and won't give issues.
🔶 SETTINGS
Anchor 1 / Anchor 2: both points will form the Training Window .
Forecasting Length: Forecasting horizon, determines how many bars in the 'future' are forecasted.
Smooth: Controls the degree of smoothness of the model fit.
Sigma: Noise variance. Controls the amplitude of the forecast, lower values will make it more sensitive to outliers.
Machine Learning: VWAP [YinYangAlgorithms]Machine Learning: VWAP aims to use Machine Learning to Identify the best location to Anchor the VWAP at. Rather than using a traditional fixed length or simply adjusting based on a Date / Time; by applying Machine Learning we may hope to identify crucial areas which make sense to reset the VWAP and start anew. VWAP’s may act similar to a Bollinger Band in the sense that they help to identify both Overbought and Oversold Price locations based on previous movements and help to identify how far the price may move within the current Trend. However, unlike Bollinger Bands, VWAPs have the ability to parabolically get quite spaced out and also reset. For this reason, the price may never actually go from the Lower to the Upper and vice versa (when very spaced out; when the Upper and Lower zones are narrow, it may bounce between the two). The reason for this is due to how the anchor location is calculated and in this specific Indicator, how it changes anchors based on price movement calculated within Machine Learning.
This Indicator changes the anchor if the Low < Lowest Low of a length of X and likewise if the High > Highest High of a length of X. This logic is applied within a Machine Learning standpoint that likewise amplifies this Lookback Length by adding a Machine Learning Length to it and increasing the lookback length even further.
Due to how the anchor for this VWAP changes, you may notice that the Basis Line (Orange) may act as a Trend Identifier. When the Price is above the basis line, it may represent a bullish trend; and likewise it may represent a bearish trend when below it. You may also notice what may happen is when the trend occurs, it may push all the way to the Upper or Lower levels of this VWAP. It may then proceed to move horizontally until the VWAP expands more and it may gain more movement; or it may correct back to the Basis Line. If it corrects back to the basis line, what may happen is it either uses the Basis Line as a Support and continues in its current direction, or it will change the VWAP anchor and start anew.
Tutorial:
If we zoom in on the most recent VWAP we can see how it expands. Expansion may be caused by time but generally it may be caused by price movement and volume. Exponential Price movement causes the VWAP to expand, even if there are corrections to it. However, please note Volume adds a large weighted factor to the calculation; hence Volume Weighted Average Price (VWAP).
If you refer to the white circle in the example above; you’ll be able to see that the VWAP expanded even while the price was correcting to the Basis line. This happens due to exponential movement which holds high volume. If you look at the volume below the white circle, you’ll notice it was very large; however even though there was exponential price movement after the white circle, since the volume was low, the VWAP didn’t expand much more than it already had.
There may be times where both Volume and Price movement isn’t significant enough to cause much of an expansion. During this time it may be considered to be in a state of consolidation. While looking at this example, you may also notice the color switch from red to green to red. The color of the VWAP is related to the movement of the Basis line (Orange middle line). When the current basis is > the basis of the previous bar the color of the VWAP is green, and when the current basis is < the basis of the previous bar, the color of the VWAP is red. The color may help you gauge the current directional movement the price is facing within the VWAP.
You may have noticed there are signals within this Indicator. These signals are composed of Green and Red Triangles which represent potential Bullish and Bearish momentum changes. The Momentum changes happen when the Signal Type:
The High/Low or Close (You pick in settings)
Crosses one of the locations within the VWAP.
Bullish Momentum change signals occur when :
Signal Type crosses OVER the Basis
Signal Type crosses OVER the lower level
Bearish Momentum change signals occur when:
Signal Type crosses UNDER the Basis
Signal Type Crosses UNDER the upper level
These signals may represent locations where momentum may occur in the direction of these signals. For these reasons there are also alerts available to be set up for them.
If you refer to the two circles within the example above, you may see that when the close goes above the basis line, how it mat represents bullish momentum. Likewise if it corrects back to the basis and the basis acts as a support, it may continue its bullish momentum back to the upper levels again. However, if you refer to the red circle, you’ll see if the basis fails to act as a support, it may then start to correct all the way to the lower levels, or depending on how expanded the VWAP is, it may just reset its anchor due to such drastic movement.
You also have the ability to disable Machine Learning by setting ‘Machine Learning Type’ to ‘None’. If this is done, it will go off whether you have it set to:
Bullish
Bearish
Neutral
For the type of VWAP you want to see. In this example above we have it set to ‘Bullish’. Non Machine Learning VWAP are still calculated using the same logic of if low < lowest low over length of X and if high > highest high over length of X.
Non Machine Learning VWAP’s change much quicker but may also allow the price to correct from one side to the other without changing VWAP Anchor. They may be useful for breaking up a trend into smaller pieces after momentum may have changed.
Above is an example of how the Non Machine Learning VWAP looks like when in Bearish. As you can see based on if it is Bullish or Bearish is how it favors the trend to be and may likewise dictate when it changes the Anchor.
When set to neutral however, the Anchor may change quite quickly. This results in a still useful VWAP to help dictate possible zones that the price may move within, but they’re also much tighter zones that may not expand the same way.
We will conclude this Tutorial here, hopefully this gives you some insight as to why and how Machine Learning VWAPs may be useful; as well as how to use them.
Settings:
VWAP:
VWAP Type: Type of VWAP. You can favor specific direction changes or let it be Neutral where there is even weight to both. Please note, these do not apply to the Machine Learning VWAP.
Source: VWAP Source. By default VWAP usually uses HLC3; however OHLC4 may help by providing more data.
Lookback Length: The Length of this VWAP when it comes to seeing if the current High > Highest of this length; or if the current Low is < Lowest of this length.
Standard VWAP Multiplier: This multiplier is applied only to the Standard VWMA. This is when 'Machine Learning Type' is set to 'None'.
Machine Learning:
Use Rational Quadratics: Rationalizing our source may be beneficial for usage within ML calculations.
Signal Type: Bullish and Bearish Signals are when the price crosses over/under the basis, as well as the Upper and Lower levels. These may act as indicators to where price movement may occur.
Machine Learning Type: Are we using a Simple ML Average, KNN Mean Average, KNN Exponential Average or None?
KNN Distance Type: We need to check if distance is within the KNN Min/Max distance, which distance checks are we using.
Machine Learning Length: How far back is our Machine Learning going to keep data for.
k-Nearest Neighbour (KNN) Length: How many k-Nearest Neighbours will we account for?
Fast ML Data Length: What is our Fast ML Length? This is used with our Slow Length to create our KNN Distance.
Slow ML Data Length: What is our Slow ML Length? This is used with our Fast Length to create our KNN Distance.
If you have any questions, comments, ideas or concerns please don't hesitate to contact us.
HAPPY TRADING!
Machine Learning: Optimal RSI [YinYangAlgorithms]This Indicator, will rate multiple different lengths of RSIs to determine which RSI to RSI MA cross produced the highest profit within the lookback span. This ‘Optimal RSI’ is then passed back, and if toggled will then be thrown into a Machine Learning calculation. You have the option to Filter RSI and RSI MA’s within the Machine Learning calculation. What this does is, only other Optimal RSI’s which are in the same bullish or bearish direction (is the RSI above or below the RSI MA) will be added to the calculation.
You can either (by default) use a Simple Average; which is essentially just a Mean of all the Optimal RSI’s with a length of Machine Learning. Or, you can opt to use a k-Nearest Neighbour (KNN) calculation which takes a Fast and Slow Speed. We essentially turn the Optimal RSI into a MA with different lengths and then compare the distance between the two within our KNN Function.
RSI may very well be one of the most used Indicators for identifying crucial Overbought and Oversold locations. Not only that but when it crosses its Moving Average (MA) line it may also indicate good locations to Buy and Sell. Many traders simply use the RSI with the standard length (14), however, does that mean this is the best length?
By using the length of the top performing RSI and then applying some Machine Learning logic to it, we hope to create what may be a more accurate, smooth, optimal, RSI.
Tutorial:
This is a pretty zoomed out Perspective of what the Indicator looks like with its default settings (except with Bollinger Bands and Signals disabled). If you look at the Tables above, you’ll notice, currently the Top Performing RSI Length is 13 with an Optimal Profit % of: 1.00054973. On its default settings, what it does is Scan X amount of RSI Lengths and checks for when the RSI and RSI MA cross each other. It then records the profitability of each cross to identify which length produced the overall highest crossing profitability. Whichever length produces the highest profit is then the RSI length that is used in the plots, until another length takes its place. This may result in what we deem to be the ‘Optimal RSI’ as it is an adaptive RSI which changes based on performance.
In our next example, we changed the ‘Optimal RSI Type’ from ‘All Crossings’ to ‘Extremity Crossings’. If you compare the last two examples to each other, you’ll notice some similarities, but overall they’re quite different. The reason why is, the Optimal RSI is calculated differently. When using ‘All Crossings’ everytime the RSI and RSI MA cross, we evaluate it for profit (short and long). However, with ‘Extremity Crossings’, we only evaluate it when the RSI crosses over the RSI MA and RSI <= 40 or RSI crosses under the RSI MA and RSI >= 60. We conclude the crossing when it crosses back on its opposite of the extremity, and that is how it finds its Optimal RSI.
The way we determine the Optimal RSI is crucial to calculating which length is currently optimal.
In this next example we have zoomed in a bit, and have the full default settings on. Now we have signals (which you can set alerts for), for when the RSI and RSI MA cross (green is bullish and red is bearish). We also have our Optimal RSI Bollinger Bands enabled here too. These bands allow you to see where there may be Support and Resistance within the RSI at levels that aren’t static; such as 30 and 70. The length the RSI Bollinger Bands use is the Optimal RSI Length, allowing it to likewise change in correlation to the Optimal RSI.
In the example above, we’ve zoomed out as far as the Optimal RSI Bollinger Bands go. You’ll notice, the Bollinger Bands may act as Support and Resistance locations within and outside of the RSI Mid zone (30-70). In the next example we will highlight these areas so they may be easier to see.
Circled above, you may see how many times the Optimal RSI faced Support and Resistance locations on the Bollinger Bands. These Bollinger Bands may give a second location for Support and Resistance. The key Support and Resistance may still be the 30/50/70, however the Bollinger Bands allows us to have a more adaptive, moving form of Support and Resistance. This helps to show where it may ‘bounce’ if it surpasses any of the static levels (30/50/70).
Due to the fact that this Indicator may take a long time to execute and it can throw errors for such, we have added a Setting called: Adjust Optimal RSI Lookback and RSI Count. This settings will automatically modify the Optimal RSI Lookback Length and the RSI Count based on the Time Frame you are on and the Bar Indexes that are within. For instance, if we switch to the 1 Hour Time Frame, it will adjust the length from 200->90 and RSI Count from 30->20. If this wasn’t adjusted, the Indicator would Timeout.
You may however, change the Setting ‘Adjust Optimal RSI Lookback and RSI Count’ to ‘Manual’ from ‘Auto’. This will give you control over the ‘Optimal RSI Lookback Length’ and ‘RSI Count’ within the Settings. Please note, it will likely take some “fine tuning” to find working settings without the Indicator timing out, but there are definitely times you can find better settings than our ‘Auto’ will create; especially on higher Time Frames. The Minimum our ‘Auto’ will create is:
Optimal RSI Lookback Length: 90
RSI Count: 20
The Maximum it will create is:
Optimal RSI Lookback Length: 200
RSI Count: 30
If there isn’t much bar index history, for instance, if you’re on the 1 Day and the pair is BTC/USDT you’ll get < 4000 Bar Indexes worth of data. For this reason it is possible to manually increase the settings to say:
Optimal RSI Lookback Length: 500
RSI Count: 50
But, please note, if you make it too high, it may also lead to inaccuracies.
We will conclude our Tutorial here, hopefully this has given you some insight as to how calculating our Optimal RSI and then using it within Machine Learning may create a more adaptive RSI.
Settings:
Optimal RSI:
Show Crossing Signals: Display signals where the RSI and RSI Cross.
Show Tables: Display Information Tables to show information like, Optimal RSI Length, Best Profit, New Optimal RSI Lookback Length and New RSI Count.
Show Bollinger Bands: Show RSI Bollinger Bands. These bands work like the TDI Indicator, except its length changes as it uses the current RSI Optimal Length.
Optimal RSI Type: This is how we calculate our Optimal RSI. Do we use all RSI and RSI MA Crossings or just when it crosses within the Extremities.
Adjust Optimal RSI Lookback and RSI Count: Auto means the script will automatically adjust the Optimal RSI Lookback Length and RSI Count based on the current Time Frame and Bar Index's on chart. This will attempt to stop the script from 'Taking too long to Execute'. Manual means you have full control of the Optimal RSI Lookback Length and RSI Count.
Optimal RSI Lookback Length: How far back are we looking to see which RSI length is optimal? Please note the more bars the lower this needs to be. For instance with BTC/USDT you can use 500 here on 1D but only 200 for 15 Minutes; otherwise it will timeout.
RSI Count: How many lengths are we checking? For instance, if our 'RSI Minimum Length' is 4 and this is 30, the valid RSI lengths we check is 4-34.
RSI Minimum Length: What is the RSI length we start our scans at? We are capped with RSI Count otherwise it will cause the Indicator to timeout, so we don't want to waste any processing power on irrelevant lengths.
RSI MA Length: What length are we using to calculate the optimal RSI cross' and likewise plot our RSI MA with?
Extremity Crossings RSI Backup Length: When there is no Optimal RSI (if using Extremity Crossings), which RSI should we use instead?
Machine Learning:
Use Rational Quadratics: Rationalizing our Close may be beneficial for usage within ML calculations.
Filter RSI and RSI MA: Should we filter the RSI's before usage in ML calculations? Essentially should we only use RSI data that are of the same type as our Optimal RSI? For instance if our Optimal RSI is Bullish (RSI > RSI MA), should we only use ML RSI's that are likewise bullish?
Machine Learning Type: Are we using a Simple ML Average, KNN Mean Average, KNN Exponential Average or None?
KNN Distance Type: We need to check if distance is within the KNN Min/Max distance, which distance checks are we using.
Machine Learning Length: How far back is our Machine Learning going to keep data for.
k-Nearest Neighbour (KNN) Length: How many k-Nearest Neighbours will we account for?
Fast ML Data Length: What is our Fast ML Length? This is used with our Slow Length to create our KNN Distance.
Slow ML Data Length: What is our Slow ML Length? This is used with our Fast Length to create our KNN Distance.
If you have any questions, comments, ideas or concerns please don't hesitate to contact us.
HAPPY TRADING!
Machine Learning: SuperTrend Strategy TP/SL [YinYangAlgorithms]The SuperTrend is a very useful Indicator to display when trends have shifted based on the Average True Range (ATR). Its underlying ideology is to calculate the ATR using a fixed length and then multiply it by a factor to calculate the SuperTrend +/-. When the close crosses the SuperTrend it changes direction.
This Strategy features the Traditional SuperTrend Calculations with Machine Learning (ML) and Take Profit / Stop Loss applied to it. Using ML on the SuperTrend allows for the ability to sort data from previous SuperTrend calculations. We can filter the data so only previous SuperTrends that follow the same direction and are within the distance bounds of our k-Nearest Neighbour (KNN) will be added and then averaged. This average can either be achieved using a Mean or with an Exponential calculation which puts added weight on the initial source. Take Profits and Stop Losses are then added to the ML SuperTrend so it may capitalize on Momentum changes meanwhile remaining in the Trend during consolidation.
By applying Machine Learning logic and adding a Take Profit and Stop Loss to the Traditional SuperTrend, we may enhance its underlying calculations with potential to withhold the trend better. The main purpose of this Strategy is to minimize losses and false trend changes while maximizing gains. This may be achieved by quick reversals of trends where strategic small losses are taken before a large trend occurs with hopes of potentially occurring large gain. Due to this logic, the Win/Loss ratio of this Strategy may be quite poor as it may take many small marginal losses where there is consolidation. However, it may also take large gains and capitalize on strong momentum movements.
Tutorial:
In this example above, we can get an idea of what the default settings may achieve when there is momentum. It focuses on attempting to hit the Trailing Take Profit which moves in accord with the SuperTrend just with a multiplier added. When momentum occurs it helps push the SuperTrend within it, which on its own may act as a smaller Trailing Take Profit of its own accord.
We’ve highlighted some key points from the last example to better emphasize how it works. As you can see, the White Circle is where profit was taken from the ML SuperTrend simply from it attempting to switch to a Bullish (Buy) Trend. However, that was rejected almost immediately and we went back to our Bearish (Sell) Trend that ended up resulting in our Take Profit being hit (Yellow Circle). This Strategy aims to not only capitalize on the small profits from SuperTrend to SuperTrend but to also capitalize when the Momentum is so strong that the price moves X% away from the SuperTrend and is able to hit the Take Profit location. This Take Profit addition to this Strategy is crucial as momentum may change state shortly after such drastic price movements; and if we were to simply wait for it to come back to the SuperTrend, we may lose out on lots of potential profit.
If you refer to the Yellow Circle in this example, you’ll notice what was talked about in the Summary/Overview above. During periods of consolidation when there is little momentum and price movement and we don’t have any Stop Loss activated, you may see ‘Signal Flashing’. Signal Flashing is when there are Buy and Sell signals that keep switching back and forth. During this time you may be taking small losses. This is a normal part of this Strategy. When a signal has finally been confirmed by Momentum, is when this Strategy shines and may produce the profit you desire.
You may be wondering, what causes these jagged like patterns in the SuperTrend? It's due to the ML logic, and it may be a little confusing, but essentially what is happening is the Fast Moving SuperTrend and the Slow Moving SuperTrend are creating KNN Min and Max distances that are extreme due to (usually) parabolic movement. This causes fewer values to be added to and averaged within the ML and causes less smooth and more exponential drastic movements. This is completely normal, and one of the perks of using k-Nearest Neighbor for ML calculations. If you don’t know, the Min and Max Distance allowed is derived from the most recent(0 index of data array) to KNN Length. So only SuperTrend values that exhibit distances within these Min/Max will be allowed into the average.
Since the KNN ML logic can cause these exponential movements in the SuperTrend, they likewise affect its Take Profit. The Take Profit may benefit from this movement like displayed in the example above which helped it claim profit before then exhibiting upwards movement.
By default our Stop Loss Multiplier is kept quite low at 0.0000025. Keeping it low may help to reduce some Signal Flashing while not taking extra losses more so than not using it at all. However, if we increase it even more to say 0.005 like is shown in the example above. It can really help the trend keep momentum. Please note, although previous results don’t imply future results, at 0.0000025 Stop Loss we are currently exhibiting 69.27% profit while at 0.005 Stop Loss we are exhibiting 33.54% profit. This just goes to show that although there may be less Signal Flashing, it may not result in more profit.
We will conclude our Tutorial here. Hopefully this has given you some insight as to how Machine Learning, combined with Trailing Take Profit and Stop Loss may have positive effects on the SuperTrend when turned into a Strategy.
Settings:
SuperTrend:
ATR Length: ATR Length used to create the Original Supertrend.
Factor: Multiplier used to create the Original Supertrend.
Stop Loss Multiplier: 0 = Don't use Stop Loss. Stop loss can be useful for helping to prevent false signals but also may result in more loss when hit and less profit when switching trends.
Take Profit Multiplier: Take Profits can be useful within the Supertrend Strategy to stop the price reverting all the way to the Stop Loss once it's been profitable.
Machine Learning:
Only Factor Same Trend Direction: Very useful for ensuring that data used in KNN is not manipulated by different SuperTrend Directional data. Please note, it doesn't affect KNN Exponential.
Rationalized Source Type: Should we Rationalize only a specific source, All or None?
Machine Learning Type: Are we using a Simple ML Average, KNN Mean Average, KNN Exponential Average or None?
Machine Learning Smoothing Type: How should we smooth our Fast and Slow ML Datas to be used in our KNN Distance calculation? SMA, EMA or VWMA?
KNN Distance Type: We need to check if distance is within the KNN Min/Max distance, which distance checks are we using.
Machine Learning Length: How far back is our Machine Learning going to keep data for.
k-Nearest Neighbour (KNN) Length: How many k-Nearest Neighbours will we account for?
Fast ML Data Length: What is our Fast ML Length?? This is used with our Slow Length to create our KNN Distance.
Slow ML Data Length: What is our Slow ML Length?? This is used with our Fast Length to create our KNN Distance.
If you have any questions, comments, ideas or concerns please don't hesitate to contact us.
HAPPY TRADING!
Machine Learning using Neural Networks | EducationalThe script provided is a comprehensive illustration of how to implement and execute a simplistic Neural Network (NN) on TradingView using PineScript.
It encompasses the entire workflow from data input, weight initialization, implicit neuron calculation, feedforward computation, backpropagation for weight adjustments, generating predictions, to visualizing the Mean Squared Error (MSE) Loss Curve for monitoring the training phase.
In the visual example above, you can see that the prediction is not aligned with the actual value. This is intentional for demonstrative purposes, and by incrementing the Epochs or Learning Rate, you will see these two values converge as the accuracy increases.
Hyperparameters:
Learning Rate, Epochs, and the choice between Simple Backpropagation and a verbose version are declared as script inputs, allowing users to tailor the training process.
Initialization:
Random initialization of weight matrices (w1, w2) is performed to ensure asymmetry, promoting effective gradient updates. A seed is added for reproducibility.
Utility Functions:
Functions for matrix randomization, sigmoid activation, MSE loss calculation, data normalization, and standardization are defined to streamline the computation process.
Neural Network Computation:
The feedforward function computes the hidden and output layer values given the input.
Two variants of the backpropagation function are provided for weight adjustment, with one offering a more verbose step-by-step computation of gradients.
A wrapper train_nn function iterates through epochs, performing feedforward, loss computation, and backpropagation in each epoch while logging and collecting loss values.
Training Invocation:
The input data is prepared by normalizing it to a value between 0 and 1 using the maximum standardized value, and the training process is invoked only on the last confirmed bar to preserve computational resources.
Output Forecasting and Visualization:
Post training, the NN's output (predicted price) is computed, standardized and visualized alongside the actual price on the chart.
The MSE loss between the predicted and actual prices is visualized, providing insight into the prediction accuracy.
Optionally, the MSE Loss Curve is plotted on the chart, illustrating the loss trajectory through epochs, assisting in understanding the training performance.
Customizable Visualization:
Various inputs control visualization aspects like Chart Scaling, Chart Horizontal Offset, and Chart Vertical Offset, allowing users to adapt the visualization to their preference.
-------------------------------------------------------
The following is this Neural Network structure, consisting of one hidden layer, with two hidden neurons.
Through understanding the steps outlined in my code, one should be able to scale the NN in any way they like, such as changing the input / output data and layers to fit their strategy ideas.
Additionally, one could forgo the backpropagation function, and load their own trained weights into the w1 and w2 matrices, to have this code run purely for inference.
-------------------------------------------------------
While this demonstration does create a “prediction”, it is on historical data. The purpose here is educational, rather than providing a ready tool for non-programmer consumers.
Normally in Machine Learning projects, the training process would be split into two segments, the Training and the Validation parts. For the purpose of conveying the core concept in a concise and non-repetitive way, I have foregone the Validation part. However, it is merely the application of your trained network on new data (feedforward), and monitoring the loss curve.
Essentially, checking the accuracy on “unseen” data, while training it on “seen” data.
-------------------------------------------------------
I hope that this code will help developers create interesting machine learning applications within the Tradingview ecosystem.
Machine Learning: Donchian DCA Grid Strategy [YinYangAlgorithms]This strategy uses a Machine Learning approach on the Donchian Channels with a DCA and Grid purchase/sell Strategy. Not only that, but it uses a custom Bollinger calculation to determine its Basis which is used as a mild sell location. This strategy is a pure DCA strategy in the sense that no shorts are used and theoretically it can be used in webhooks on most exchanges as it’s only using Spot Orders. The idea behind this strategy is we utilize both the Highest Highs and Lowest Lows within a Machine Learning standpoint to create Buy and Sell zones. We then fraction these zones off into pieces to create Grids. This allows us to ‘micro’ purchase as it enters these zones and likewise ‘micro’ sell as it goes up into the upper (sell) zones.
You have the option to set how many grids are used, by default we use 100 with max 1000. These grids can be ‘stacked’ together if a single bar is to go through multiple at the same time. For instance, if a bar goes through 30 grids in one bar, it will have a buy/sell power of 30x. Stacking Grid Buy and (sometimes) Sells is a very crucial part of this strategy that allows it to purchase multitudes during crashes and capitalize on sales during massive pumps.
With the grids, you’ll notice there is a middle line within the upper and lower part that makes the grid. As a Purchase Type within our Settings this is identified as ‘Middle of Zone Purchase Amount In USDT’. The middle of the grid may act as the strongest grid location (aside from maybe the bottom). Therefore there is a specific purchase amount for this Grid location.
This DCA Strategy also features two other purchase methods. Most importantly is its ‘Purchase More’ type. Essentially it will attempt to purchase when the Highest High or Lowest Low moves outside of the Outer band. For instance, the Lowest Low becomes Lower or the Higher High becomes Higher. When this happens may be a good time to buy as it is featuring a new High or Low over an extended period.
The last but not least Purchase type within this Strategy is what we call a ‘Strong Buy’. The reason for this is its verified by the following:
The outer bounds have been pushed (what causes a ‘Purchase More’)
The Price has crossed over the EMA 21
It has been verified through MACD, RSI or MACD Historical (Delta) using Regular and Hidden Divergence (Note, only 1 of these verifications is required and it can be any).
By default we don’t have Purchase Amount for ‘Strong Buy’ set, but that doesn’t mean it can’t be viable, it simply means we have only seen a few pairs where it actually proved more profitable allocating money there rather than just increasing the purchase amount for ‘Purchase More’ or ‘Grids’.
Now that you understand where we BUY, we should discuss when we SELL.
This Strategy features 3 crucial sell locations, and we will discuss each individually as they are very important.
1. ‘Sell Some At’: Here there are 4 different options, by default its set to ‘Both’ but you can change it around if you want. Your options are:
‘Both’ - You will sell some at both locations. The amount sold is the % used at ‘Sell Some %’.
‘Basis Line’ - You will sell some when the price crosses over the Basis Line. The amount sold is the % used at ‘Sell Some %’.
‘Percent’ - You will sell some when the Close is >= X% between the Lower Inner and Upper Inner Zone.
‘None’ - This simply means don’t ever Sell Some.
2. Sell Grids. Sell Grids are exactly like purchase grids and feature the same amount of grids. You also have the ability to ‘Stack Grid Sells’, which basically means if a bar moves multiple grids, it will stack the amount % wise you will sell, rather than just selling the default amount. Sell Grids use a DCA logic but for selling, which we deem may help adjust risk/reward ratio for selling, especially if there is slow but consistent bullish movement. It causes these grids to constantly push up and therefore when the close is greater than them, accrue more profit.
3. Take Profit. Take profit occurs when the close first goes above the Take Profit location (Teal Line) and then Closes below it. When Take Profit occurs, ALL POSITIONS WILL BE SOLD. What may happen is the price enters the Sell Grid, doesn’t go all the way to the top ‘Exiting it’ and then crashes back down and closes below the Take Profit. Take Profit is a strong location which generally represents a strong profit location, and that a strong momentum has changed which may cause the price to revert back to the buy grid zone.
Keep in mind, if you have (by default) ‘Only Sell If Profit’ toggled, all sell locations will only create sell orders when it is profitable to do so. Just cause it may be a good time to sell, doesn’t mean based on your DCA it is. In our opinion, only selling when it is profitable to do so is a key part of the DCA purchase strategy.
You likewise have the ability to ‘Only Buy If Lower than DCA’, which is likewise by default. These two help keep the Yin and Yang by balancing each other out where you’re only purchasing and selling when it makes logical sense too, even if that involves ignoring a signal and waiting for a better opportunity.
Tutorial:
Like most of our Strategies, we try to capitalize on lower Time Frames, generally the 15 minutes so we may find optimal entry and exit locations while still maintaining a strong correlation to trend patterns.
First off, let’s discuss examples of how this Strategy works prior to applying Machine Learning (enabled by default).
In this example above we have disabled the showing of ‘Potential Buy and Sell Signals’ so as to declutter the example. In here you can see where actual trades had gone through for both buying and selling and get an idea of how the strategy works. We also have disabled Machine Learning for this example so you can see the hard lines created by the Donchian Channel. You can also see how the Basis line ‘white line’ may act as a good location to ‘Sell Some’ and that it moves quite irregularly compared to the Donchian Channel. This is due to the fact that it is based on two custom Bollinger Bands to create the basis line.
Here we zoomed out even further and moved back a bit to where there were dense clusters of buy and sell orders. Sometimes when the price is rather volatile you’ll see it ‘Ping Pong’ back and forth between the buy and sell zones quite quickly. This may be very good for your trades and profit as a whole, especially if ‘Only Buy If Lower Than DCA’ and ‘Only Sell If Profit’ are both enabled; as these toggles will ensure you are:
Always lowering your Average when buying
Always making profit when selling
By default 8% commission is added to the Strategy as well, to simulate the cost effects of if these trades were taking place on an actual exchange.
In this example we also turned on the visuals for our ‘Purchase More’ (orange line) and ‘Take Profit’ (teal line) locations. These are crucial locations. The Purchase More makes purchases when the bottom of the grid has been moved (may dictate strong price movement has occurred and may be potential for correction). Our Take Profit may help secure profit when a momentum change is happening and all of the Sell Grids weren’t able to be used.
In the example above we’ve enabled Buy and Sell Signals so that you can see where the Take Profit and Purchase More signals have occurred. The white circle demonstrates that not all of the Position Size was sold within the Sell Grids, and therefore it was ALL CLOSED when the price closed below the Take Profit Line (Teal).
Then, when the bottom of the Donchian Channel was pushed further down due to the close (within the yellow circle), a Purchase More Signal was triggered.
When the close keeps pushing the bottom of the Buy Grid lower, it can cause multiple Purchase More Signals to occur. This is normal and also a crucial part of this strategy to help lower your DCA. Please note, the Purchase More won’t trigger a Buy if the Close is greater than the DCA and you have ‘Only Purchase If Lower Than DCA’ activated.
By turning on Machine Learning (default settings) the Buy and Sell Grid Zones are smoothed out more. It may cause it to look quite a bit different. Machine Learning although it looks much worse, may help increase the profit this Strategy can produce. Previous results DO NOT mean future results, but in this example, prior to turning on Machine Learning it had produced 37% Profit in ~5 months and with Machine Learning activated it is now up to 57% Profit in ~5 months.
Machine Learning causes the Strategy to focus less on Grids and more on Purchase More when it comes to getting its entries. However, if you likewise attempt to focus on Purchase More within non Machine Learning, the locations are different and therefore the results may not be as profitable.
PLEASE NOTE:
By default this strategy uses 1,000,000 as its initial capital. The amount it purchases in its Settings is relevant to this Initial capital. Considering this is a DCA Strategy, we only want to ‘Micro’ Buy and ‘Micro’ Sell whenever conditions are met.
Therefore, if you increase the Initial Capital, you’ll likewise want to increase the Purchase Amounts within the Settings and Vice Versa. For instance, if you wish to set the Initial Capital to 10,000, you should likewise can the amounts in the Settings to 1% of what they are to account for this.
We may change the Purchase Amounts to be based on %’s in a later update if it is requested.
We will conclude this Tutorial here, hopefully you can see how a DCA Grid Purchase Model applied to Machine Learning Donchian Channels may be useful for making strategic purchases in low and high zones.
Settings:
Display Data:
Show Potential Buy Locations: These locations are where 'Potentially' orders can be placed. Placement of orders is dependant on if you have 'Only Buy If Lower Than DCA' toggled and the Price is lower than DCA. It also is effected by if you actually have any money left to purchase with; you can't buy if you have no money left!
Show Potential Sell Locations: These locations are where 'Potentially' orders will be sold. If 'Only Sell If Profit' is toggled, the sell will only happen if you'll make profit from it!
Show Grid Locations: Displaying won't affect your trades but it can be useful to see where trades will be placed, as well as which have gone through and which are left to be purchased. Max 100 Grids, but visuals will only be shown if its 20 or less.
Purchase Settings:
Only Buy if its lower than DCA: Generally speaking, we want to lower our Average, and therefore it makes sense to only buy when the close is lower than our current DCA and a Purchase Condition is met.
Compound Purchases: Compounding Purchases means reinvesting profit back into your trades right away. It drastically increases profits, but it also increases risk too. It will adjust your Purchase Amounts for the Purchase Type you have set at the same % rate of strategy initial_capital to the amounts you have set.
Adjust Purchase Amount Ratio to Maintain Risk level: By adjusting purchase levels we generally help maintain a safe risk level. Basically we generally want to reserve X amount of % for each purchase type being used and relocate money when there is too much in one type. This helps balance out purchase amounts and ensure the types selected have a correct ratio to ensure they can place the right amount of orders.
Stack Grid Buys: Stacking Buy Grids is when the Close crosses multiple Buy Grids within the same bar. Should we still only purchase the value of 1 Buy Grid OR stack the grid buys based on how many buy grids it went through.
Purchase Type: Where do you want to make Purchases? We recommend lowering your risk by combining All purchase types, but you may also customize your trading strategy however you wish.
Strong Buy Purchase Amount In USDT: How much do you want to purchase when the 'Strong Buy' signal appears? This signal only occurs after it has at least entered the Buy Zone and there have been other verifications saying it's now a good time to buy. Our Strong Buy Signal is a very strong indicator that a large price movement towards the Sell Zone will likely occur. It almost always results in it leaving the Buy Zone and usually will go to at least the White Basis line where you can 'Sell Some'.
Buy More Purchase Amount In USDT: How much should you purchase when the 'Purchase More' signal appears? This 'Purchase More' signal occurs when the lowest level of the Buy Zone moves lower. This is a great time to buy as you're buying the dip and generally there is a correction that will allow you to 'Sell Some' for some profit.
Amount of Grid Buy and Sells: How many Grid Purchases do you want to make? We recommend having it at the max of 10, as it will essentially get you a better Average Purchase Price, but you may adjust it to whatever you wish. This amount also only matters if your Purchase Type above incorporates Grid Purchases. Max 100 Grids, but visuals will only be shown if it's 20 or less.
Each Grid Purchase Amount In USDT: How much should you purchase after closing under a grid location? Keep in mind, if you have 10 grids and it goes through each, it will be this amount * 10. Grid purchasing is a great way to get a good entry, lower risk and also lower your average.
Middle Of Zone Purchase Amount In USDT: The Middle Of Zone is the strongest grid location within the Buy Zone. This is why we have a unique Purchase Amount for this Grid specifically. Please note you need to have 'Middle of Zone is a Grid' enabled for this Purchase Amount to be used.
Sell:
Only Sell if its Profit: There is a chance that during a dump, all your grid buys when through, and a few Purchase More Signals have appeared. You likely got a good entry. A Strong Buy may also appear before it starts to pump to the Sell Zone. The issue that may occur is your Average Purchase Price is greater than the 'Sell Some' price and/or the Grids in the Sell Zone and/or the Strong Sell Signal. When this happens, you can either take a loss and sell it, or you can hold on to it and wait for more purchase signals to therefore lower your average more so you can take profit at the next sell location. Please backtest this yourself within our YinYang Purchase Strategy on the pair and timeframe you are wanting to trade on. Please also note, that previous results will not always reflect future results. Please assess the risk yourself. Don't trade what you can't afford to lose. Sometimes it is better to strategically take a loss and continue on making profit than to stay in a bad trade for a long period of time.
Stack Grid Sells: Stacking Sell Grids is when the Close crosses multiple Sell Grids within the same bar. Should we still only sell the value of 1 Sell Grid OR stack the grid sells based on how many sell grids it went through.
Stop Loss Type: This is when the Close has pushed the Bottom of the Buy Grid More. Do we Stop Loss or Purchase More?? By default we recommend you stay true to the DCA part of this strategy by Purchasing More, but this is up to you.
Sell Some At: Where if selected should we 'Sell Some', this may be an important way to sell a little bit at a good time before the price may correct. Also, we don't want to sell too much incase it doesn't correct though, so its a 'Sell Some' location. Basis Line refers to our Moving Basis Line created from 2 Bollinger Bands and Percent refers to a Percent difference between the Lower Inner and Upper Inner bands.
Sell Some At Percent Amount: This refers to how much % between the Lower Inner and Upper Inner bands we should well at if we chose to 'Sell Some'.
Sell Some Min %: This refers to the Minimum amount between the Lower Inner band and Close that qualifies a 'Sell Some'. This acts as a failsafe so we don't 'Sell Some' for too little.
Sell % At Strong Sell Signal: How much do we sell at the 'Strong Sell' Signal? It may act as a strong location to sell, but likewise Grid Sells could be better.
Grid and Donchian Settings:
Donchian Channel Length: How far back are we looking back to determine our Donchian Channel.
Extra Outer Buy Width %: How much extra should we push the Outer Buy (Low) Width by?
Extra Inner Buy Width %: How much extra should we push the Inner Buy (Low) Width by?
Extra Inner Sell Width %: How much extra should we push the Inner Sell (High) Width by?
Extra Outer Sell Width %: How much extra should we push the Outer Sell (High) Width by?
Machine Learning:
Rationalized Source Type: Donchians usually use High/Low. What Source is our Rationalized Source using?
Machine Learning Type: Are we using a Simple ML Average, KNN Mean Average, KNN Exponential Average or None?
Machine Learning Length: How far back is our Machine Learning going to keep data for.
k-Nearest Neighbour (KNN) Length: How many k-Nearest Neighbours will we account for?
Fast ML Data Length: What is our Fast ML Length?? This is used with our Slow Length to create our KNN Distance.
Slow ML Data Length: What is our Slow ML Length?? This is used with our Fast Length to create our KNN Distance.
If you have any questions, comments, ideas or concerns please don't hesitate to contact us.
HAPPY TRADING!
Machine Learning: Trend Lines [YinYangAlgorithms]Trend lines have always been a key indicator that may help predict many different types of price movements. They have been well known to create different types of formations such as: Pennants, Channels, Flags and Wedges. The type of formation they create is based on how the formation was created and the angle it was created. For instance, if there was a strong price increase and then there is a Wedge where both end points meet, this is considered a Bull Pennant. The formations Trend Lines create may be powerful tools that can help predict current Support and Resistance and also Future Momentum changes. However, not all Trend Lines will create formations, and alone they may stand as strong Support and Resistance locations on the Vertical.
The purpose of this Indicator is to apply Machine Learning logic to a Traditional Trend Line Calculation, and therefore allowing a new approach to a modern indicator of high usage. The results of such are quite interesting and goes to show the impacts a simple KNN Machine Learning model can have on Traditional Indicators.
Tutorial:
There are a few different settings within this Indicator. Many will greatly impact the results and if any are changed, lots will need ‘Fine Tuning’. So let's discuss the main toggles that have great effects and what they do before discussing the lengths. Currently in this example above we have the Indicator at its Default Settings. In this example, you can see how the Trend Lines act as key Support and Resistance locations. Due note, Support and Resistance are a relative term, as is their color. What starts off as Support or Resistance may change when the price crosses over / under them.
In the example above we have zoomed in and circled locations that exhibited markers of Support and Resistance along the Trend Lines. These Trend Lines are all created using the Default Settings. As you can see from the example above; just because it is a Green Upwards Trend Line, doesn’t mean it’s a Support Line. Support and Resistance is always shifting on Trend Lines based on the prices location relative to them.
We won’t go through all the Formations Trend Lines make, but the example above, we can see the Trend Lines formed a Downward Channel. Channels are when there are two parallel downwards Trend Lines that are at a relatively similar angle. This means that they won’t ever meet. What may happen when the price is within these channels, is it may bounce between the upper and lower bounds. These Channels may drive the price upwards or downwards, depending on if it is in an Upwards or Downwards Channel.
If you refer to the example above, you’ll notice that the Trend Lines are formed like traditional Trend Lines. They don’t stem from current Highs and Lows but rather Machine Learning Highs and Lows. More often than not, the Machine Learning approach to Trend Lines cause their start point and angle to be quite different than a Traditional Trend Line. Due to this, it may help predict Support and Resistance locations at are more uncommon and therefore can be quite useful.
In the example above we have turned off the toggle in Settings ‘Use Exponential Data Average’. This Settings uses a custom Exponential Data Average of the KNN rather than simply averaging the KNN. By Default it is enabled, but as you can see when it is disabled it may create some pretty strong lasting Trend Lines. This is why we advise you ZOOM OUT AS FAR AS YOU CAN. Trend Lines are only displayed when you’ve zoomed out far enough that their Start Point is visible.
As you can see in this example above, there were 3 major Upward Trend Lines created in 2020 that have had a major impact on Support and Resistance Locations within the last year. Lets zoom in and get a closer look.
We have zoomed in for this example above, and circled some of the major Support and Resistance locations that these Upward Trend Lines may have had a major impact on.
Please note, these Machine Learning Trend Lines aren’t a ‘One Size Fits All’ kind of thing. They are completely customizable within the Settings, so that you can get a tailored experience based on what Pair and Time Frame you are trading on.
When any values are changed within the Settings, you’ll likely need to ‘Fine Tune’ the rest of the settings until your desired result is met. By default the modifiable lengths within the Settings are:
Machine Learning Length: 50
KNN Length:5
Fast ML Data Length: 5
Slow ML Data Length: 30
For example, let's toggle ‘Use Exponential Data Averages’ back on and change ‘Fast ML Data Length’ from 5 to 20 and ‘Slow ML Data Length’ from 30 to 50.
As you can in the example above, all of the lines have changed. Although there are still some strong Support Locations created by the Upwards Trend Lines.
We will conclude our Tutorial here. Hopefully you’ve learned how to use Machine Learning Trend Lines and will be able to now see some more unorthodox Support and Resistance locations on the Vertical.
Settings:
Use Machine Learning Sources: If disabled Traditional Trend line sources (High and Low) will be used rather than Rational Quadratics.
Use KNN Distance Sorting: You can disable this if you wish to not have the Machine Learning Data sorted using KNN. If disabled trend line logic will be Traditional.
Use Exponential Data Average: This Settings uses a custom Exponential Data Average of the KNN rather than simply averaging the KNN.
Machine Learning Length: How strong is our Machine Learning Memory? Please note, when this value is too high the data is almost 'too' much and can lead to poor results.
K-Nearest Neighbour (KNN) Length: How many K-Nearest Neighbours are allowed with our Distance Clustering? Please note, too high or too low may lead to poor results.
Fast ML Data Length: Fast and Slow speed needs to be adjusted properly to see results. 3/5/7 all seem to work well for Fast.
Slow ML Data Length: Fast and Slow speed needs to be adjusted properly to see results. 20 - 50 all seem to work well for Slow.
If you have any questions, comments, ideas or concerns please don't hesitate to contact us.
HAPPY TRADING!
Machine Learning: MFI Heat Map [YinYangAlgorithms]Overview:
MFI Heat Maps are a visually appealing way to display the values of 29 different MFIs at the same time while being able to make sense of it. Each plot within the Indicator represents a different MFI value. The higher you get up, the longer the length that was used for this MFI. This Indicator also features the use of Machine Learning to help balance the MFI levels. It doesn’t solely rely upon Machine Learning but instead incorporates a growing length MFI averaged with the Machine Learning MFI at any given index.
For instance, say we are calculating the 10th plot from the bottom, the MFI would be an average of:
MFI(source, 11)
Machine Learning MFI at Index of 10
We do it this way as they both help smooth each other out without relying solely on just one calculation method.
Due to plot limitations, you are capped at 28 Plot Amounts within this indicator, but that is still quite a bit of information you can glean from a Heat Map.
The Machine Learning used in this indicator is of the K-Nearest Neighbor (KNN). It uses a Fast and Slow MFI calculation then sorts through them over Machine Learning Length and calculates the differences between them. It then slices off KNN length to create our Max/Min Distances allotted. It adds the average between Fast and Slow MFIs to a Viable Distances array if their distances are within the KNN Min/Max distance. It then averages all distances in the Viable Distances array and returns the result.
The result of the KNN Function is saved to another ML Data array whose length is that of Plot Amount (Heat Map Size). This way each Index of the ML Data array can be indexed according to the Heat Map Size.
The Average of the ML Data array is the MFI line (white) that you’ll see plotted on the Indicator. There is also the SMA of the MFI Average (orange) which is likewise plotted. These plots allow you to visualize where the ML MFI is sitting and can potentially be useful for seeing when the MFI Average and SMA cross over and under each other.
We’ve heard many people talk highly of RSI, but sadly not too many even refer to MFI. MFI oftentimes may be overlooked, especially with new traders who may not even know what it is. Essentially MFI is an RSI but it also incorporates Volume into its calculations, which in our opinion leads to a more accurate reading; afterall, what is price movement without Volume.
Tutorial:
You may be thinking, this Indicator looks appealing to the eye, but how do I benefit from it trading wise?
Before we get into our visual examples, let's talk briefly about what makes Heat Maps in general a useful tool for trading. Heat Maps give us the ability to visualize and understand lots of data while removing the clutter. We can understand the data of 29 different MFIs without having to look at and decipher 29 different MFI plots. When you overlay too many MFI lines on top of each other, they can be very difficult to read and oftentimes end up actually hindering your Technical Analysis. For this reason, we have a simple solution to this problem; Heat Maps. This MFI Heat Map allows you to easily know (in a relative %) what the MFI level is for varying lengths. For Instance, the First (bottom) plot indexes an MFI of (K(0) (loop of Plot Amount) + Smoothing Length (default 1)) = 1. Since this is indexing (usually) a very low length, it will change much quicker. Whereas the Last (top) plot indexes an MFI of (K(27) (loop of Plot Amount) + Smoothing Length (default 1)) = 28. This is indexing a much higher length of MFI which results in the MFI the higher you go up in the Heat Map to move much slower.
Heat Maps give us the ability to see changes happening over multiple MFIs at the same time, which can be very useful for seeing shifts in MFI / Momentum. Remember, MFI incorporates Volume, so even if the price goes up a lot, if there was low volume, the MFI won’t move as much as an RSI would. However, likewise, if there is high volume but low price movement, the MFI will move slightly more than the RSI.
Heat Maps change color based on their MFI level. If the MFI is >= 90 it is HOT (red), if the MFI <= 9 it is COLD (teal, think of ICE). Green represents an MFI of 50-59 and Dark Blue represents an MFI of 40-49. Green and Dark blue are the most common colors as all the others are more ‘Extreme’ MFI levels.
Okay, time to get to the Examples :
Since there is so much going on in Heat Maps, we’ve decided to focus this tutorial to this specific area and talk about individual locations before talking about it as a whole.
If you refer to the example above where there are 2 white circles; these white circles are highlighting a key location you’ll be wanting to identify within your Heat Maps, many things are happening here:
The MFI crossed over the SMA (bullish).
The Heat Map started changing from mid/dark Blue (30-50 MFI) to Green (50-59 MFI) around the midline (the 50% dashed like).
The Lower levels of the Heat Map are turning Yellow/Orange/Red (60-100 MFI).
The Upper Levels of the Heat Map are still Light Blue - Green (10-50 MFI).
The 4 Key points above, all point towards potential Bullish Momentum changes. You’re likely wondering, but why? Let's discuss about each one in more specific detail:
1. The MFI crossed over the SMA (bullish): What this tells us is that the current MFI Average is now greater than its average over the last (default) 16 bars. This means there's been a large amount of Money Flow (Price and Volume) recently (subjectively based on the last (default) 16 average). This is one of the leading Bullish / Bearish signals you will see within this Indicator. You can enable Signals within the Settings and/or even add Alerts for when these crossings occur.
2. The Heat Map started changing from mid/dark Blue (30-50 MFI) to Green (50-59 MFI) around the midline (the 50% dashed like): This shows us that the index’s in the mid (if using all 28 heat map plots it would be at 14) has already received some of this momentum change. If you look at the second white circle (right), you’ll also notice the higher MFI plot indexes are also green. This is because since their length is long they still have some momentum and strength from the first white circle (left). Just because the first white circle failed in its bullish push, doesn’t mean it didn’t achieve momentum that would later on help to push the price up.
3. The Lower levels of the Heat Map are turning Yellow/Orange/Red (60-100 MFI): It occurred somewhat in the left white circle, but mainly in the right white circle. This shows us the MFI is very high on the lower lengths, this may lead to the current, middle and higher length MFIs following suit soon. Remember it has to work its way up, the higher levels can’t go red unless the lower levels go red first and the higher levels can also lag quite a bit behind and take awhile to catch up, this is normal, expected and meant to happen. Vice versa is also true with getting higher levels to go cold (light teal (think of ICE)).
4. The Upper Levels of the Heat Map are still Light Blue - Green (10-50 MFI): You might think at first that this is a bad thing, but it's not! Remember you want to be Fearful when others are Greedy and Greedy when others are Fearful! You don’t want to buy when the higher levels have a high MFI, you want to buy when you see the momentum pushing up in the lower MFI levels (getting yellow/orange/red in the low levels) while it is still Cold in the higher levels (BLUE OR GREEN, nothing higher than green as it is already slightly too high). There will be many times that it is Yellow or possibly Orange in the high levels and the bullish push still happens, but this is much more risky! The key to trading is to minimize risks while maximizing potential.
Hopefully now you’re getting an idea of how to spot potential bullish momentum changes, but what about bearish momentum changes? Technically they are the exact opposite, so we don’t need to go into as much detail, but lets still take a look at a few examples:
In the example above we marked the 3 times where it was displaying overly bullish characteristics. We marked the bullish momentum occurring with arrows. If you look closely at the start of the arrow to where it finishes, you’ll notice how the heat (HOT)(RED) works its way up from the lower levels to the higher levels. We then see the MFI to SMA cross under. In all 3 of these examples the heat made it all the way to the top of the chart. These are all very bearish signals that represent a bearish momentum movement that may occur soon.
Also, please note, the level the MFI is at DOES matter! That line isn’t there simply for you to see when there are crosses over and under. The MFI is considered to be Overbought when it is greater than 70 (the upper white dashed line, it is just formatted to be on a different scale cause there are 28 plots, but it represents 70). The MFI is considered to be Oversold when it is less than 30 (the lower white dashed line).
If we look to the left a little here where a big drop in price occurred shortly after our MFI and SMA crossed, would we have been able to identify it using the Heat Maps? Likely, No. There was some color change in the lower levels a few bars prior that went yellow/orange/red but before this cross happened they all went back to Dark Blue. In the middle section when the cross happened it was only Green and Yellow and in the upper section we are Blue. This would be a very risky trade to go on as the only real Bearish Indication was the MFI to SMA cross under. Remember, you want to reduce risk, you don’t want to simply trade on everytime the MFI and SMA cross each other or you’ll be getting yourself into many risky trades based on false signals.
Based on what you’ve learned above, can you see the signs that are indicating where this white circle may have potential for a bullish momentum change?
Now that we are more zoomed in, you may also be noticing there are colors to the price bars. This can be disabled in the settings, but just so you know what they mean, let’s zoom in a little more and talk about it.
We’ve condensed the Indicator a bit so you can see the bars better here. The colors that are displayed on these bars are the Heat Map value for your MFI (the white line in the Indicator). This way you can better see when the Price is Hot and Cold. As you may see while looking, the colors generally go from cold to hot when bullish momentum is happening and hot to cold when bearish momentum is happening. We don’t recommend solely looking at the bars as indicators to MFI momentum change, as seeing the Heat Map will give you much more data; however it can be nice to see the Heat Map projected on the bars rather than trying to eyeball it yourself or hover over each bar specifically to see their levels.
We will conclude our Tutorial here. Hopefully this has given you some insight to how useful Heat Maps can be and why it works well with a Machine Learning (KNN) Model applied to the MFI.
PLEASE NOTE: You can adjust the line width for the Heat Map within the settings. If you condense the Indicator a lot or have a small screen, likely use a length of 1-2. If you have it stretched out or a large screen, a length of 2-3 will work nice. You just don’t want to have the lines overlapping or it defeats the purpose of a Heat Map. Also, the bigger the linewidth, generally you’ll want to increase the Transparency within the Settings also as it can get quite bright and hurt your eyes over time.
Settings:
MFI:
Show MFI and SMA Crossing Signals: MFI and SMA Crossing is one of the leading Bullish and Bearish Signals in this Indicator. You can also add alerts for these signals.
Plot Amount: How many plots are used in this Heat Map. (2 - 28).
Source: The Source to use in all MFI calculations.
Smooth Initial MFI Length: How much to smooth the Fast and Slow MFI calculation by. 1 = No smoothing.
MFI SMA Length: What length we smooth the MFI Average over to get our MFI SMA.
Machine Learning:
Average MFI data by adding a lookback to the Source: While populating our Heat Map with the MFI's, should use use the Source each MFI Length increase or should we also lookback a Source each MFI Length Increase.
KNN Distance Requirement: To be a valid KNN, it needs to abide by a Distance calculation. Generally only Max is used, but you can change it if it suits your trading style better.
Machine Learning Length: How much ML data should we store? The longer the length generally the smoother the result; which may not be as accurate for something like a Heat Map, so keeping this relatively low may lead to more accurate results.
KNN Length: How many KNN are used in the slice to calculate max/min distance allowed.
Fast Length: Fast MFI length used in KNN to calculate distances by comparing its distance with the Slow MFI Length.
Slow Length: Slow MFI length used in KNN to calculate distances by comparing its distance with the Fast MFI Length.
Smoothing Length: When populating our Heat Map, at what length do we start our MFI calculations with (A Higher value with result in a slower and more smoothed MFI / Heat Map).
Colors:
Change Bar Color: Change bar colors to MFI Avg Color.
Heat Map Transparency: If there isn't any transparency it can be a little hard on the eyes. The Greater the Line Width, generally the more transparency you'll want for your eyes.
Line Width: Set how wide the Heat Map lines are
MFI 90-100 Color: Color when the MFI is between these levels.
MFI 80-89 Color: Color when the MFI is between these levels.
MFI 70-79 Color: Color when the MFI is between these levels.
MFI 60-69 Color: Color when the MFI is between these levels.
MFI 50-59 Color: Color when the MFI is between these levels.
MFI 40-49 Color: Color when the MFI is between these levels.
MFI 30-39 Color: Color when the MFI is between these levels.
MFI 20-29 Color: Color when the MFI is between these levels.
MFI 10-19 Color: Color when the MFI is between these levels.
MFI 0-100 Color: Color when the MFI is between these levels.
If you have any questions, comments, ideas or concerns please don't hesitate to contact us.
HAPPY TRADING!
AI Momentum [YinYang]Overview:
AI Momentum is a kernel function based momentum Indicator. It uses Rational Quadratics to help smooth out the Moving Averages, this may give them a more accurate result. This Indicator has 2 main uses, first it displays ‘Zones’ that help you visualize the potential movement areas and when the price is out of bounds (Overvalued or Undervalued). Secondly it creates signals that display the momentum of the current trend.
The Zones are composed of the Highest Highs and Lowest lows turned into a Rational Quadratic over varying lengths. These create our Rational High and Low zones. There is however a second zone. The second zone is composed of the avg of the Inner High and Inner Low zones (yellow line) and the Rational Quadratic of the current Close. This helps to create a second zone that is within the High and Low bounds that may represent momentum changes within these zones. When the Rationalized Close crosses above the High and Low Zone Average it may signify a bullish momentum change and vice versa when it crosses below.
There are 3 different signals created to display momentum:
Bullish and Bearish Momentum. These signals display when there is current bullish or bearish momentum happening within the trend. When the momentum changes there will likely be a lull where there are neither Bullish or Bearish momentum signals. These signals may be useful to help visualize when the momentum has started and stopped for both the bulls and the bears. Bullish Momentum is calculated by checking if the Rational Quadratic Close > Rational Quadratic of the Highest OHLC4 smoothed over a VWMA. The Bearish Momentum is calculated by checking the opposite.
Overly Bullish and Bearish Momentum. These signals occur when the bar has Bullish or Bearish Momentum and also has an Rationalized RSI greater or less than a certain level. Bullish is >= 57 and Bearish is <= 43. There is also the option to ‘Factor Volume’ into these signals. This means, the Overly Bullish and Bearish Signals will only occur when the Rationalized Volume > VWMA Rationalized Volume as well as the previously mentioned factors above. This can be useful for removing ‘clutter’ as volume may dictate when these momentum changes will occur, but it can also remove some of the useful signals and you may miss the swing too if the volume just was low. Overly Bullish and Bearish Momentum may dictate when a momentum change will occur. Remember, they are OVERLY Bullish and Bearish, meaning there is a chance a correction may occur around these signals.
Bull and Bear Crosses. These signals occur when the Rationalized Close crosses the Gaussian Close that is 2 bars back. These signals may show when there is a strong change in momentum, but be careful as more often than not they’re predicting that the momentum may change in the opposite direction.
Tutorial:
As we can see in the example above, generally what happens is we get the regular Bullish or Bearish momentum, followed by the Rationalized Close crossing the Zone average and finally the Overly Bullish or Bearish signals. This is normally the order of operations but isn’t always how it happens as sometimes momentum changes don’t make it that far; also the Rationalized Close and Zone Average don’t follow any of the same math as the Signals which can result in differing appearances. The Bull and Bear Crosses are also quite sporadic in appearance and don’t generally follow any sort of order of operations. However, they may occur as a Predictor between Bullish and Bearish momentum, signifying the beginning of the momentum change.
The Bull and Bear crosses may be a Predictor of momentum change. They generally happen when there is no Bullish or Bearish momentum happening; and this helps to add strength to their prediction. When they occur during momentum (orange circle) there is a less likely chance that it will happen, and may instead signify the exact opposite; it may help predict a large spike in momentum in the direction of the Bullish or Bearish momentum. In the case of the orange circle, there is currently Bearish Momentum and therefore the Bull Cross may help predict a large momentum movement is about to occur in favor of the Bears.
We have disabled signals here to properly display and talk about the zones. As you can see, Rationalizing the Highest Highs and Lowest Lows over 2 different lengths creates inner and outer bounds that help to predict where parabolic movement and momentum may move to. Our Inner and Outer zones are great for seeing potential Support and Resistance locations.
The secondary zone, which can cross over and change from Green to Red is also a very important zone. Let's zoom in and talk about it specifically.
The Middle Zone Crosses may help deduce where parabolic movement and strong momentum changes may occur. Generally what may happen is when the cross occurs, you will see parabolic movement to the High / Low zones. This may be the Inner zone but can sometimes be the outer zone too. The hard part is sometimes it can be a Fakeout, like displayed with the Blue Circle. The Cross doesn’t mean it may move to the opposing side, sometimes it may just be predicting Parabolic movement in a general sense.
When we turn the Momentum Signals back on, we can see where the Fakeout occurred that it not only almost hit the Inner Low Zone but it also exhibited 2 Overly Bearish Signals. Remember, Overly bearish signals mean a momentum change in favor of the Bulls may occur soon and overly Bullish signals mean a momentum change in favor of the Bears may occur soon.
You may be wondering, well what does “may occur soon” mean and how do we tell?
The purpose of the momentum signals is not only to let you know when Momentum has occurred and when it is still prevalent. It also matters A LOT when it has STOPPED!
In this example above, we look at when the Overly Bullish and Bearish Momentum has STOPPED. As you can see, when the Overly Bullish or Bearish Momentum stopped may be a strong predictor of potential momentum change in the opposing direction.
We will conclude our Tutorial here, hopefully this Indicator has been helpful for showing you where momentum is occurring and help predict how far it may move. We have been dabbling with and are planning on releasing a Strategy based on this Indicator shortly.
Settings:
1. Momentum:
Show Signals: Sometimes it can be difficult to visualize the zones with signals enabled.
Factor Volume: Factor Volume only applies to Overly Bullish and Bearish Signals. It's when the Volume is > VWMA Volume over the Smoothing Length.
Zone Inside Length: The Zone Inside is the Inner zone of the High and Low. This is the length used to create it.
Zone Outside Length: The Zone Outside is the Outer zone of the High and Low. This is the length used to create it.
Smoothing length: Smoothing length is the length used to smooth out our Bullish and Bearish signals, along with our Overly Bullish and Overly Bearish Signals.
2. Kernel Settings:
Lookback Window: The number of bars used for the estimation. This is a sliding value that represents the most recent historical bars. Recommended range: 3-50.
Relative Weighting: Relative weighting of time frames. As this value approaches zero, the longer time frames will exert more influence on the estimation. As this value approaches infinity, the behavior of the Rational Quadratic Kernel will become identical to the Gaussian kernel. Recommended range: 0.25-25.
Start Regression at Bar: Bar index on which to start regression. The first bars of a chart are often highly volatile, and omission of these initial bars often leads to a better overall fit. Recommended range: 5-25.
If you have any questions, comments, ideas or concerns please don't hesitate to contact us.
HAPPY TRADING!
AI Channels (Clustering) [LuxAlgo]The AI Channels indicator is constructed based on rolling K-means clustering, a common machine learning method used for clustering analysis. These channels allow users to determine the direction of the underlying trends in the price.
We also included an option to display the indicator as a trailing stop from within the settings.
🔶 USAGE
Each channel extremity allows users to determine the current trend direction. Price breaking over the upper extremity suggesting an uptrend, and price breaking below the lower extremity suggesting a downtrend. Using a higher Window Size value will return longer-term indications.
The "Clusters" setting allows users to control how easy it is for the price to break an extremity, with higher values returning extremities further away from the price.
The "Denoise Channels" is enabled by default and allows to see less noisy extremities that are more coherent with the detected trend.
Users who wish to have more focus on a detected trend can display the indicator as a trailing stop.
🔹 Centroid Dispersion Areas
Each extremity is made of one area. The width of each area indicates how spread values within a cluster are around their centroids. A wider area would suggest that prices within a cluster are more spread out around their centroid, as such one could say that it is indicative of the volatility of a cluster.
Wider areas around a specific extremity can indicate a larger and more spread-out amount of prices within the associated cluster. In practice price entering an area has a higher chance to break an associated extremity.
🔶 DETAILS
The indicator performs K-means clustering over the most recent Window Size prices, finding a number of user-specified clusters. See here to find more information on cluster detection.
The channel extremities are returned as the centroid of the lowest, average, and highest price clusters.
K-means clustering can be computationally expensive and as such we allow users to determine the maximum number of iterations used to find the centroids as well as the number of most historical bars to perform the indicator calculation. Do note that increasing the calculation window of the indicator as well as the number of clusters will return slower results.
🔶 SETTINGS
Window Size: Amount of most recent prices to use for the calculation of the indicator.
Clusters": Amount of clusters detected for the calculation of the indicator.
Denoise Channels: When enabled, return less noisy channels extremities, disabling this setting will return the exact centroids at each time but will produce less regular extremities.
As Trailing Stop: Display the indicator as a trailing stop.
🔹 Optimization
This group of settings affects the runtime performance of the script.
Maximum Iteration Steps: Maximum number of iterations allowed for finding centroids. Excessively low values can return a better script load time but poor clustering.
Historical Bars Calculation: Calculation window of the script (in bars).
Machine Learning Regression Trend [LuxAlgo]The Machine Learning Regression Trend tool uses random sample consensus (RANSAC) to fit and extrapolate a linear model by discarding potential outliers, resulting in a more robust fit.
🔶 USAGE
The proposed tool can be used like a regular linear regression, providing support/resistance as well as forecasting an estimated underlying trend.
Using RANSAC allows filtering out outliers from the input data of our final fit, by outliers we are referring to values deviating from the underlying trend whose influence on a fitted model is undesired. For financial prices and under the assumptions of segmented linear trends, these outliers can be caused by volatile moves and/or periodic variations within an underlying trend.
Adjusting the "Allowed Error" numerical setting will determine how sensitive the model is to outliers, with higher values returning a more sensitive model. The blue margin displayed shows the allowed error area.
The number of outliers in the calculation window (represented by red dots) can also be indicative of the amount of noise added to an underlying linear trend in the price, with more outliers suggesting more noise.
Compared to a regular linear regression which does not discriminate against any point in the calculation window, we see that the model using RANSAC is more conservative, giving more importance to detecting a higher number of inliners.
🔶 DETAILS
RANSAC is a general approach to fitting more robust models in the presence of outliers in a dataset and as such does not limit itself to a linear regression model.
This iterative approach can be summarized as follow for the case of our script:
Step 1: Obtain a subset of our dataset by randomly selecting 2 unique samples
Step 2: Fit a linear regression to our subset
Step 3: Get the error between the value within our dataset and the fitted model at time t , if the absolute error is lower than our tolerance threshold then that value is an inlier
Step 4: If the amount of detected inliers is greater than a user-set amount save the model
Repeat steps 1 to 4 until the set number of iterations is reached and use the model that maximizes the number of inliers
🔶 SETTINGS
Length: Calculation window of the linear regression.
Width: Linear regression channel width.
Source: Input data for the linear regression calculation.
🔹 RANSAC
Minimum Inliers: Minimum number of inliers required to return an appropriate model.
Allowed Error: Determine the tolerance threshold used to detect potential inliers. "Auto" will automatically determine the tolerance threshold and will allow the user to multiply it through the numerical input setting at the side. "Fixed" will use the user-set value as the tolerance threshold.
Maximum Iterations Steps: Maximum number of allowed iterations.
Pro Supertrend CalculatorThis indicator is an adapted version of Julien_Eche's 'Pro Momentum Calculator' tailored specifically for TradingView's 'Supertrend indicator'.
The "Pro Supertrend Calculator" indicator has been developed to provide traders with a data-driven perspective on price movements in financial markets. Its primary objective is to analyze historical price data and make probabilistic predictions about the future direction of price movements, specifically in terms of whether the next candlestick will be bullish (green) or bearish (red). Here's a deeper technical insight into how it accomplishes this task:
1. Supertrend Computation:
The indicator initiates by computing the Supertrend indicator, a sophisticated technical analysis tool. This calculation involves two essential parameters:
- ATR Length (Average True Range Length): This parameter determines the sensitivity of the Supertrend to price fluctuations.
- Factor: This multiplier plays a pivotal role in establishing the distance between the Supertrend line and prevailing market prices. A higher factor value results in a more significant separation.
2. Supertrend Visualization:
The Supertrend values derived from the calculation are meticulously plotted on the price chart, manifesting as two distinct lines:
- Green Line: This line represents the Supertrend when it indicates a bullish trend, signifying an anticipation of rising prices.
- Red Line: This line signifies the Supertrend in bearish market conditions, indicating an expectation of falling prices.
3. Consecutive Candle Analysis:
- The core function of the indicator revolves around tracking successive candlestick patterns concerning their relationship with the Supertrend line.
- To be included in the analysis, a candlestick must consistently close either above (green candles) or below (red candles) the Supertrend line for multiple consecutive periods.
4.Labeling and Enumeration:
- To communicate the count of consecutive candles displaying uniform trend behavior, the indicator meticulously applies labels to the price chart.
- The positioning of these labels varies based on the direction of the trend, residing either below (for bullish patterns) or above (for bearish patterns) the candlestick.
- The color scheme employed aligns with the color of the candle, using green labels for bullish candles and red labels for bearish ones.
5. Tabular Data Presentation:
- The indicator augments its graphical analysis with a customizable table prominently displayed on the chart. This table delivers comprehensive statistical insights.
- The tabular data comprises the following key elements for each consecutive period:
a. Consecutive Candles: A tally of the number of consecutive candles displaying identical trend characteristics.
b. Candles Above Supertrend: A count of candles that remained above the Supertrend during the sequential period.
3. Candles Below Supertrend: A count of candles that remained below the Supertrend during the sequential period.
4. Upcoming Green Candle: An estimation of the probability that the next candlestick will be bullish, grounded in historical data.
5. Upcoming Red Candle: An estimation of the probability that the next candlestick will be bearish, based on historical data.
6. Tailored Configuration:
To accommodate diverse trading strategies and preferences, the indicator offers extensive customization options. Traders can fine-tune parameters such as ATR length, factor, label and table placement, and table size to align with their unique trading approaches.
In summation, the "Pro Supertrend Calculator" indicator is an intricately designed tool that leverages the Supertrend indicator in conjunction with historical price data to furnish traders with an informed outlook on potential future price dynamics, with a particular emphasis on the likelihood of specific bullish or bearish candlestick patterns stemming from consecutive price behavior.
Support & Resistance AI (K means/median) [ThinkLogicAI]█ OVERVIEW
K-means is a clustering algorithm commonly used in machine learning to group data points into distinct clusters based on their similarities. While K-means is not typically used directly for identifying support and resistance levels in financial markets, it can serve as a tool in a broader analysis approach.
Support and resistance levels are price levels in financial markets where the price tends to react or reverse. Support is a level where the price tends to stop falling and might start to rise, while resistance is a level where the price tends to stop rising and might start to fall. Traders and analysts often look for these levels as they can provide insights into potential price movements and trading opportunities.
█ BACKGROUND
The K-means algorithm has been around since the late 1950s, making it more than six decades old. The algorithm was introduced by Stuart Lloyd in his 1957 research paper "Least squares quantization in PCM" for telecommunications applications. However, it wasn't widely known or recognized until James MacQueen's 1967 paper "Some Methods for Classification and Analysis of Multivariate Observations," where he formalized the algorithm and referred to it as the "K-means" clustering method.
So, while K-means has been around for a considerable amount of time, it continues to be a widely used and influential algorithm in the fields of machine learning, data analysis, and pattern recognition due to its simplicity and effectiveness in clustering tasks.
█ COMPARE AND CONTRAST SUPPORT AND RESISTANCE METHODS
1) K-means Approach:
Cluster Formation: After applying the K-means algorithm to historical price change data and visualizing the resulting clusters, traders can identify distinct regions on the price chart where clusters are formed. Each cluster represents a group of similar price change patterns.
Cluster Analysis: Analyze the clusters to identify areas where clusters tend to form. These areas might correspond to regions of price behavior that repeat over time and could be indicative of support and resistance levels.
Potential Support and Resistance Levels: Based on the identified areas of cluster formation, traders can consider these regions as potential support and resistance levels. A cluster forming at a specific price level could suggest that this level has been historically significant, causing similar price behavior in the past.
Cluster Standard Deviation: In addition to looking at the means (centroids) of the clusters, traders can also calculate the standard deviation of price changes within each cluster. Standard deviation is a measure of the dispersion or volatility of data points around the mean. A higher standard deviation indicates greater price volatility within a cluster.
Low Standard Deviation: If a cluster has a low standard deviation, it suggests that prices within that cluster are relatively stable and less likely to exhibit sudden and large price movements. Traders might consider placing tighter stop-loss orders for trades within these clusters.
High Standard Deviation: Conversely, if a cluster has a high standard deviation, it indicates greater price volatility within that cluster. Traders might opt for wider stop-loss orders to allow for potential price fluctuations without getting stopped out prematurely.
Cluster Density: Each data point is assigned to a cluster so a cluster that is more dense will act more like gravity and
2) Traditional Approach:
Trendlines: Draw trendlines connecting significant highs or lows on a price chart to identify potential support and resistance levels.
Chart Patterns: Identify chart patterns like double tops, double bottoms, head and shoulders, and triangles that often indicate potential reversal points.
Moving Averages: Use moving averages to identify levels where the price might find support or resistance based on the average price over a specific period.
Psychological Levels: Identify round numbers or levels that traders often pay attention to, which can act as support and resistance.
Previous Highs and Lows: Identify significant previous price highs and lows that might act as support or resistance.
The key difference lies in the approach and the foundation of these methods. Traditional methods are based on well-established principles of technical analysis and market psychology, while the K-means approach involves clustering price behavior without necessarily incorporating market sentiment or specific price patterns.
It's important to note that while the K-means approach might provide an interesting way to analyze price data, it should be used cautiously and in conjunction with other traditional methods. Financial markets are influenced by a wide range of factors beyond just price behavior, and the effectiveness of any method for identifying support and resistance levels should be thoroughly tested and validated. Additionally, developments in trading strategies and analysis techniques could have occurred since my last update.
█ K MEANS ALGORITHM
The algorithm for K means is as follows:
Initialize cluster centers
assign data to clusters based on minimum distance
calculate cluster center by taking the average or median of the clusters
repeat steps 1-3 until cluster centers stop moving
█ LIMITATIONS OF K MEANS
There are 3 main limitations of this algorithm:
Sensitive to Initializations: K-means is sensitive to the initial placement of centroids. Different initializations can lead to different cluster assignments and final results.
Assumption of Equal Sizes and Variances: K-means assumes that clusters have roughly equal sizes and spherical shapes. This may not hold true for all types of data. It can struggle with identifying clusters with uneven densities, sizes, or shapes.
Impact of Outliers: K-means is sensitive to outliers, as a single outlier can significantly affect the position of cluster centroids. Outliers can lead to the creation of spurious clusters or distortion of the true cluster structure.
█ LIMITATIONS IN APPLICATION OF K MEANS IN TRADING
Trading data often exhibits characteristics that can pose challenges when applying indicators and analysis techniques. Here's how the limitations of outliers, varying scales, and unequal variance can impact the use of indicators in trading:
Outliers are data points that significantly deviate from the rest of the dataset. In trading, outliers can represent extreme price movements caused by rare events, news, or market anomalies. Outliers can have a significant impact on trading indicators and analyses:
Indicator Distortion: Outliers can skew the calculations of indicators, leading to misleading signals. For instance, a single extreme price spike could cause indicators like moving averages or RSI (Relative Strength Index) to give false signals.
Risk Management: Outliers can lead to overly aggressive trading decisions if not properly accounted for. Ignoring outliers might result in unexpected losses or missed opportunities to adjust trading strategies.
Different Scales: Trading data often includes multiple indicators with varying units and scales. For example, prices are typically in dollars, volume in units traded, and oscillators have their own scale. Mixing indicators with different scales can complicate analysis:
Normalization: Indicators on different scales need to be normalized or standardized to ensure they contribute equally to the analysis. Failure to do so can lead to one indicator dominating the analysis due to its larger magnitude.
Comparability: Without normalization, it's challenging to directly compare the significance of indicators. Some indicators might have a larger numerical range and could overshadow others.
Unequal Variance: Unequal variance in trading data refers to the fact that some indicators might exhibit higher volatility than others. This can impact the interpretation of signals and the performance of trading strategies:
Volatility Adjustment: When combining indicators with varying volatility, it's essential to adjust for their relative volatilities. Failure to do so might lead to overemphasizing or underestimating the importance of certain indicators in the trading strategy.
Risk Assessment: Unequal variance can impact risk assessment. Indicators with higher volatility might lead to riskier trading decisions if not properly taken into account.
█ APPLICATION OF THIS INDICATOR
This indicator can be used in 2 ways:
1) Make a directional trade:
If a trader thinks price will go higher or lower and price is within a cluster zone, The trader can take a position and place a stop on the 1 sd band around the cluster. As one can see below, the trader can go long the green arrow and place a stop on the one standard deviation mark for that cluster below it at the red arrow. using this we can calculate a risk to reward ratio.
Calculating risk to reward: targeting a risk reward ratio of 2:1, the trader could clearly make that given that the next resistance area above that in the orange cluster exceeds this risk reward ratio.
2) Take a reversal Trade:
We can use cluster centers (support and resistance levels) to go in the opposite direction that price is currently moving in hopes of price forming a pivot and reversing off this level.
Similar to the directional trade, we can use the standard deviation of the cluster to place a stop just in case we are wrong.
In this example below we can see that shorting on the red arrow and placing a stop at the one standard deviation above this cluster would give us a profitable trade with minimal risk.
Using the cluster density table in the upper right informs the trader just how dense the cluster is. Higher density clusters will give a higher likelihood of a pivot forming at these levels and price being rejected and switching direction with a larger move.
█ FEATURES & SETTINGS
General Settings:
Number of clusters: The user can select from 3 to five clusters. A good rule of thumb is that if you are trading intraday, less is more (Think 3 rather than 5). For daily 4 to 5 clusters is good.
Cluster Method: To get around the outlier limitation of k means clustering, The median was added. This gives the user the ability to choose either k means or k median clustering. K means is the preferred method if the user things there are no large outliers, and if there appears to be large outliers or it is assumed there are then K medians is preferred.
Bars back To train on: This will be the amount of bars to include in the clustering. This number is important so that the user includes bars that are recent but not so far back that they are out of the scope of where price can be. For example the last 2 years we have been in a range on the sp500 so 505 days in this setting would be more relevant than say looking back 5 years ago because price would have to move far to get there.
Show SD Bands: Select this to show the 1 standard deviation bands around the support and resistance level or unselect this to just show the support and resistance level by itself.
Features:
Besides the support and resistance levels and standard deviation bands, this indicator gives a table in the upper right hand corner to show the density of each cluster (support and resistance level) and is color coded to the cluster line on the chart. Higher density clusters mean price has been there previously more than lower density clusters and could mean a higher likelihood of a reversal when price reaches these areas.
█ WORKS CITED
Victor Sim, "Using K-means Clustering to Create Support and Resistance", 2020, towardsdatascience.com
Chris Piech, "K means", stanford.edu
█ ACKNOLWEDGMENTS
@jdehorty- Thanks for the publish template. It made organizing my thoughts and work alot easier.
