LastAncientOne/Deep_Learning_Machine_Learning_Stock

Implements multiple regression and neural network architectures (including LSTMs for time-series forecasting) built with Python, TensorFlow, and scikit-learn to predict stock price movements using technical and fundamental analysis features. The workflow encompasses full pipeline stages—data collection, preprocessing, model training with parameter tuning, and backtesting—to evaluate both classification and regression approaches across different market timeframes. Targets investors seeking to experiment with AI-driven stock strategies while understanding model limitations like overfitting and bias-variance tradeoffs in financial prediction tasks.

1,718 stars. No commits in the last 6 months.