Stock-Prediction-Models

Includes over 30 deep learning models, such as LSTM, GRU, and attention mechanisms, detailed in the README's models section, offering a broad range of architectures for time-series forecasting experiments.

Features 23 algorithmic trading agents, from rule-based turtles to reinforcement learning variants, with implemented Jupyter notebooks like 'policy-gradient-agent.ipynb' for hands-on testing and learning.

Provides a TensorFlow.js implementation that allows users to upload historical CSV data and run stock forecasts directly in the browser, enhancing accessibility and demonstration capabilities.

Offers Monte Carlo simulations and portfolio optimization tools, as shown in the simulation notebooks, for risk assessment and strategy validation in a controlled environment.