How SFZ Capital's AI Forecasting feature uses machine learning to generate price predictions, confidence intervals, and feature importance analysis to support your trading decisions.
AI Forecasting is an optional feature available on SFZ Capital that generates machine learningâ€â€œbased market predictions. It analyses historical price data and technical indicator patterns to produce directional forecasts for stocks you're interested in.
It is designed as a supplement to your trading strategy â€- not a replacement. The predictions come with confidence intervals so you can gauge the model's certainty.
Historical price data, volume, and technical indicator values (EMA, RSI, MACD, Bollinger Bands, ATR) are gathered for the selected asset over the requested time period.
Raw data is transformed into features the ML model can learn from: indicator crossovers, momentum changes, volatility shifts, and price patterns.
A trained machine learning model processes the features and generates a directional prediction (bullish/bearish/neutral) along with a confidence score.
Each prediction includes upper and lower bounds indicating the range of likely outcomes. Wider intervals = less certainty; narrower = more conviction in the forecast.
The model reports which features (indicators, volume patterns, etc.) contributed most to the prediction, helping you understand the reasoning behind the forecast.
A clear bullish, bearish, or neutral signal based on the ML model's analysis of current market conditions relative to historical patterns.
A percentage indicating how confident the model is in its prediction. Higher confidence means more historical support for the current pattern.
A breakdown of which indicators and data points drove the prediction. Helps you understand if the signal is momentum-driven, trend-driven, or volatility-based.
AI predictions are probabilistic, not definitive. No model can predict markets with certainty. The AI Forecasting feature is intended as one input among many â€- not as a standalone trading signal. Always combine forecasts with your own strategy rules and risk management.
Before deploying a bot on a specific stock, check the AI forecast to see if the ML model's view aligns with your strategy's direction.
Use the confidence score to adjust position sizing: higher confidence forecasts could warrant larger positions within your risk limits.
Compare the feature importance output with the indicators in your bot. If the AI emphasises RSI momentum but your strategy is EMA-based, consider whether they're in agreement.