About this project

Qlib is an open-source quantitative investment platform that leverages AI technologies to support the entire investment chain, including alpha seeking, risk modeling, portfolio optimization, and order execution. It provides a comprehensive machine learning pipeline encompassing data processing, model training, and back-testing. Key capabilities include: - Support for diverse modeling paradigms: Supervised learning, market dynamics modeling, and reinforcement learning. - Integrated components: Loose-coupled modules for data management, trading strategies, executors, and analysis reports. - Automated workflows: A tool called `qrun` that automates the process of building datasets, training models, and performing evaluations. - Model Zoo: Implementation of various SOTA quant research models such as LightGBM, Transformer, Localformer, and Tabnet. - Data Infrastructure: Tools for data preparation, health checking, and automatic updates from sources like Yahoo Finance. - Online Serving: Capabilities for low-cost online model serving and automatic model rolling. Additionally, Qlib integrates with RD-Agent to support automated factor mining and model optimization in quant investment R&D.