It provides a machine learning framework for developing and evaluating high-frequency trading strategies based on limit order book dynamics.
This project provides a machine learning framework for modeling high-frequency limit order book dynamics. It includes a feature extractor for metrics like Rise Ratio and Depth Ratio, and trains various classification models (e.g., RandomForestClassifier) to predict short-term price movements. The framework evaluates strategy performance, showing outcomes and profit/loss results via figures like `prediction.png` and `P_L.png`.
It provides a machine learning framework for developing and evaluating high-frequency trading strategies based on limit order book dynamics.
Quantitative analysts, data scientists, or algorithmic traders interested in applying machine learning to high-frequency limit order book data will find this useful.