With the rapid development of artificial intelligence technology, more and more investors and researchers are leveraging AI algorithms to conduct investment research and trading strategy development. Github, as the world’s largest open source code hosting platform, contains a wealth of free AI algorithms and datasets that could facilitate ai investment research. This article summarizes some of the most useful free ai investment research resources found on github.
Qlib – AI-oriented quantitative investment platform from Microsoft
Qlib (https://github.com/microsoft/qlib) is an open source project from Microsoft, aiming to realize the potential of AI technologies in quantitative investment. It provides a full pipeline for quantitative research, including data ingestion, feature engineering, model training, backtesting and paper trading. The key features include: support for Chinese and US stock market data, 158 technical indicators, commonly used AI models like LSTM, models for reinforcement learning strategies, comprehensive documents for both Chinese and English users. Qlib makes it easy for investors to try their AI algorithms on historical data and create better quantitative investment strategies.
JoinQuant SDK – Python API for accessing JoinQuant data
JoinQuant (https://github.com/JoinQuant/jqdatasdk) is a leading quantitative trading platform in China. The jqdatasdk python package offers easy access to the commercial data services provided by JoinQuant platform. Though only free for 3 months, the data quality is considered high enough for live trading strategies. The SDK provides convenient APIs for downloading historical trading data, fundamentals data, industry classifications, trading calendar and technical indicators.
AkShare – Python library for Chinese financial data
For investors targeting the Chinese stock market, akshare (https://github.com/akfamily/akshare) is an indispensable data source. It provides trading data, fundamentals data, economic data, industry data for both Chinese stock and futures market via python interface. The data collection is based on web crawling. According to its documentation, there are over 2,800 functions for data retrieval, data cleaning and indicator calculations.
yfinance – Python library for accessing yahoo finance data
yfinance (https://github.com/ranaroussi/yfinance) provides a pythonic way to download historical market data from Yahoo Finance platform. It supports data for stocks, ETFs, mutual funds from 47 global exchanges. Some of the major features include: fetching data as pandas DataFrames, accessing extensive options data, fetching company’s fundamentals data. It could be very useful for investors targeting US stocks.
Github hosts many high-quality open source projects for AI-driven investment research, such as Microsoft Qlib for full pipeline research, JoinQuant SDK for Chinese market data, akshare for crawling Chinese data, yfinance for downloading yahoo finance data. Investors could leverage these free resources to obtain historical data, build AI models and conduct quantitative trading strategy research.