Machine Learning | Adaptive Trend Signals

The Machine Learning | Adaptive Trend Signals indicator is a comprehensive trend-following tool that utilizes K-Means clustering and Gaussian smoothing to adapt to shifting market volatility. By categorizing volatility into distinct states, it dynamically adjusts its sensitivity to provide clearer trend identification and actionable signals.

Usage

The Usage section describes how the script can be used to navigate different market environments. The indicator functions primarily as an overlay, providing a multi-layered approach to trend analysis:

• Trend Identification: Users can monitor the SmartTrend3 line (the average of two internal trend calculations) to determine the current market bias. The line changes color and transparency based on the detected trend direction and the current volatility cluster.
• Signal Interpretation: Bullish (🐮) and Bearish (🐻) markers appear when specific crossover conditions are met between the adaptive trend lines, filtered by RSI and ADX requirements. These serve as potential entry or exit points.
• Volatility Adaptation: The script automatically shifts its sensitivity based on K-Means clustering. During high volatility, the trend factor tightens to capture rapid moves, while in low volatility, it widens to avoid noise.
• Backtest Evaluation: A dashboard in the top-right corner displays the historical performance (Win Rate and Total Trades) of the underlying logic, allowing users to assess how the settings perform on a specific asset.

Details

The script integrates several advanced technical concepts to maintain accuracy across different timeframes:

• K-Means Volatility Clustering: The algorithm gathers ATR data over a 100-bar training period and iteratively groups it into three clusters: High, Medium, and Low volatility. This determines the "SmartTrend Factor" used in the primary calculations.
• Gaussian Smoothing: To remove market "jitter," a Gaussian kernel is applied to both the RSI and the final trend lines. This results in smoother transitions compared to standard moving averages.
• SmartTrend Logic: The core calculation uses a volatility-adjusted standard deviation. It requires alignment between price, RSI (>50 for bullish, <50 for bearish), and ADX (rising and >15) to confirm that a trend has sufficient momentum to be considered valid.
• Backtesting Methodology: The internal backtester uses the SmartTrend3 value for trade execution logic. It simulates a single-position system where opposite signals close existing positions.

Settings

Adaptive ML SmartTrend Settings

• ATR Length: Sets the period for the Average True Range calculation used in the K-Means volatility clustering.
• Period: The lookback period used for RSI, ADX, and standard deviation calculations.
• Base SmartTrend Factor: The base multiplier for the volatility-adjusted bands; higher values create wider, more conservative trend lines.
• SmartTrend Smoothing Length: Determines the lookback for the Gaussian kernel smoothing applied to the trend lines.

Appearance

• Show Buy/Sell Signals?: Toggles the visibility of the bullish (🐮) and bearish (🐻) signal markers.
• Bullish/Bearish Color: Customizes the colors used for trend lines and signal shapes.
• Show Backtest Table: Enables or disables the performance dashboard on the chart.
• Table Size: Adjusts the visual size of the backtest information table.

FAQ

How do I interpret the different colors on the trend line? The color indicates the trend direction (Bullish or Bearish), while the gradient/transparency reflects the current volatility cluster assigned by the machine learning algorithm.

Why do signals sometimes disappear or change on the current bar? Signals are confirmed upon bar close. Because the indicator uses adaptive smoothing and volatility clustering, the "real-time" bar may fluctuate until the data is finalized.

How can I access this indicator? You can get access on the LuxAlgo Library for charting platforms like TradingView, MetaTrader (MT4/MT5), and NinjaTrader for free.