让AlphaZero适应测试时环境变化，提升鲁棒性。
Improving Robustness of AlphaZero Algorithms to Test-Time Environment Changes
- 在标准框架上加入简单修改，增强对环境变化的适应能力。
- 即使规划预算有限，性能提升仍显著，实测有效。
- 适合部署在动态环境中的强化学习应用，如游戏、机器人。
AlphaZero框架通过结合蒙特卡洛规划与预先训练的策略-价值神经网络，实现高效决策。然而该框架通常假设训练与测试环境一致，限制了其在动态场景中的应用。本文研究了在测试环境可能发生变化的情况下部署AlphaZero智能体的问题，证明通过若干简单改进即可显著提升性能，即使在规划预算较低时亦然。实验验证了方法的有效性，代码已公开于GitHub。
原文摘要 · Abstract (English)
The AlphaZero framework provides a standard way of combining Monte Carlo planning with prior knowledge provided by a previously trained policy-value neural network. AlphaZero usually assumes that the environment on which the neural network was trained will not change at test time, which constrains its applicability. In this paper, we analyze the problem of deploying AlphaZero agents in potentially changed test environments and demonstrate how the combination of simple modifications to the standard framework can significantly boost performance, even in settings with a low planning budget available. The code is publicly available on GitHub.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。