FinRL-X: An AI-Native Modular Infrastructure for Quantitative Trading

在A股(股票)市场上训练强化学习交易智能体

This repository contains our solution to Task 1 of the FinRL Contest 2025: developing a high-performance crypto trading agent using reinforcement learning and LLM-derived market signals.

Initiate multiple trainer scripts allowing to train agents based on the FINRL library

Crypto trading bot using FinRL reinforcement learning model

Repository containing code and notebooks exploring how to build reinforcement learning agents that trade on the stock market using FinRL

Code for the published paper “Ensemble Strategy for Algorithmic Trading Using Deep Reinforcement Learning.”

Modular AI trading system using FinRL, multi-agent analysis, and real-time data pipelines.

Testing BOVA11 composition against itself using Reinforcement Learning

A progressive DRL stock trading system built on FinRL, benchmarking four model generations across VGG CNN and Transformer architectures on three capital levels. Trained on up to 50 NASDAQ tickers (2020–2025) with Historical data from Yahoo! Finance and live market data via Alpaca and real-time news sentiment scored by FinBERT and Polygon.io.

Building the strong structured code and using FinRL to find the best configurations of agent for multistock trading

RL-based stock trading with sentiment analysis using FinBERT and FinRL. Tested A2C, PPO, and TD3 on 10 US stocks against DJI benchmark.

A Deep Reinforcement Learning agent trained to solve the optimal trade execution problem, reducing costs by over 40% against industry benchmarks.

Risk-first AI trading R&D→production pipeline for WEEX: RL/ensemble strategies, anti-overfit validation (CPCV/WF), backtesting, and WEEX API execution.

Comprehensive documentation for FinRL - A Deep Reinforcement Learning Library for Financial Markets.

Risk-aware deep reinforcement learning for automated stock trading: seven DRL algorithms (A2C, PPO, DDPG, SAC, TD3, TRPO, ACKTR) with Differential-Sharpe and CVaR reward variants and a Transformer ensemble. Includes an IEEE-style paper.

Reproducible ANN vs quantum-inspired MPS signals inside a FinRL PPO trading agent