An advanced financial forecasting engine built with Streamlit and Scikit-Learn. This terminal uses a Recursive Random Forest approach to predict asset trajectories based on price-action and moving average clusters.

• Neural Forecasting: Predicts the asset price 5 days into the future.
• Self-Correction Logic: The model performs a real-time Backtest on the last 15 valid data windows. It compares its past predictions against actual historical outcomes to generate a "Confidence Score."
• Probability Zones: Uses the standard deviation of historical residuals to map an "Error Margin" cloud, visualizing market uncertainty.
• Sentiment Intelligence: Integrated AlphaVantage News API to gauge market "mood" and align it with technical indicators.

The app uses Instance-Based Learning. Every time you run a search:

1. It fetches the most recent 100 days of data.
2. It retrains the Random Forest Regressor on the fly.
3. It "Self-Corrects" by measuring the delta between its past training iterations and current reality.