/FinRL-Tutorials

Tutorials. Please star.

Primary LanguageJupyter NotebookMIT LicenseMIT

Mission: creating hundreds of user-friendly demos.

Note that we provide tutorials for FinRL-meta and FinRL.

File Structure

1-Introduction

notebooks for beginners, introduction step-by-step

  • FinRL_StockTrading_NeurIPS_2018: first tutorial notebook that trades Dow 30 using 5 DRL algorithms.
  • FinRL_PortfolioAllocation_NeurIPS_2020: provides basic settings to do portfolio allocation on Dow 30.
  • FinRL_StockTrading_Fundamental: merges fundamental indicators in earnings reports such as 'ROA', 'ROE', 'PE' with technical indicators.

2-Advance

notebooks for intermediate users

  • FinRL_PortfolioAllocation_Explainable_DRL: this notebook uses an empirical approach to explain the strategies of DRL agents for the portfolio management task. 1) it uses feature weights of a trained DRL agent, 2) histogram of correlation coefficient, 3) Z-statistics to explain the strategies.
  • FinRL_Compare_ElegantRL_RLlib_Stablebaseline3: compares popular DRL libraries, namely ElegantRL, RLlib and Stablebaseline3.
  • FinRL_Ensemble_StockTrading_ICAIF_2020: uses an ensemble strategy to combine multiple DRL agents to form an adaptive one to improve the robustness.

3-Practical

notebooks for users to explore paper trading and more financial markets

  • FinRL_PaperTrading_Demo: paper trading using FinRL through Alpaca.
  • FinRL_MultiCrypto_Trading: trading top 10 market cap cryptocurrencies.
  • FinRL_China_A_Share_Market: trading on China A Share market.

4-Optimization

notebooks for users interested in hyperparameter optimizations

5-Others

other related notebooks