Research map / Reinforcement learning / Visual and representation RL
TradeMaster: A Holistic Quantitative Trading Platform Empowered by Reinforcement Learning
Shuo Sun, Molei Qin, Wentao Zhang, Haochong Xia, Chuqiao Zong, Jie Ying, Yonggang Xie, Lingxuan Zhao, Xinrun Wang, Bo An
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use classic TD3 instead of trading-specific EIIE | portfolio management agent · DJ30 portfolio management | total return (%): 49.66 -> 57.15 | Improved |
| Use classic DDPG instead of trading-specific SARL | portfolio management agent · DJ30 portfolio management | Sharpe ratio: 0.752 -> 0.800 | Improved |
| Use EIIE instead of TD3 for risk control | portfolio management agent · DJ30 portfolio management | max drawdown (%): 32.36 -> 30.86 | Improved |
| Use EIIE instead of TD3 for volatility | portfolio management agent · DJ30 portfolio management | volatility (%): 1.373 -> 1.358 | Improved |
| Use EIIE instead of TD3 for diversity | portfolio management agent · DJ30 portfolio management | entropy: 2.608 -> 3.368 | Improved |
| Use IMIT instead of TD3 for diversity | portfolio management agent · DJ30 portfolio management | effective number of bets: 1.214 -> 2.149 | Improved |
| Use IMIT (investor imitation) instead of TD3 | portfolio management agent · DJ30 portfolio management | total return (%): 57.15 -> -4.95 | Hurt |
| Use SARL instead of TD3 for profit | portfolio management agent · DJ30 portfolio management | total return (%): 57.15 -> 52.13 | Hurt |
| Use SARL instead of TD3 for max drawdown | portfolio management agent · DJ30 portfolio management | max drawdown (%): 32.36 -> 32.23 | Improved |
| Use DDPG instead of TD3 | portfolio management agent · DJ30 portfolio management | total return (%): 57.15 -> 56.95 | No clear effect |
| Grid search over batch size, hidden size and learning rate | RL agents · DJ30 portfolio management | DJ30 performance | Improved |
| Extreme market period (Sept 2021): trading-specific EIIE/SARL vs classic RL | portfolio agents · DJ30 portfolio management | total return and Sharpe ratio (Sept 2021) | Improved |
Closest papers by meaning
- R&D-Agent-Quant: A Multi-Agent Framework for Data-Centric Factors and Model Joint Optimization (NeurIPS 2025 · similarity 0.87)
- Logic-Q: Improving Deep Reinforcement Learning-based Quantitative Trading via Program Sketch-based Tuning (AAAI 2025 · similarity 0.87)
- Robust Reinforcement Learning in Finance: Modeling Market Impact with Elliptic Uncertainty Sets (NeurIPS 2025 · similarity 0.86)
- EarnHFT: Efficient Hierarchical Reinforcement Learning for High Frequency Trading (AAAI 2024 · similarity 0.86)
- MarS: a Financial Market Simulation Engine Powered by Generative Foundation Model (ICLR 2025 · similarity 0.86)
- AlphaQCM: Alpha Discovery in Finance with Distributional Reinforcement Learning (ICML 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
