arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2006.05595cs.LGcs.AI
收藏

关系域的拟合Q-学习

Fitted Q-Learning for Relational Domains

The University of Texas at Dallas(德克萨斯大学达拉斯分校);Duke University(杜克大学);TU Darmstadt(达姆施塔特工业大学)

Srijita Das, Sriraam Natarajan, Kaushik Roy, Ronald Parr, Kristian Kersting

摘要:

研究了关系域上的近似动态规划问题。受命题环境中拟合Q-学习方法成功的启发,我们通过表示价值函数和Bellman残差,开发了第一个关系拟合Q-学习算法。当我们拟合Q函数时,我们展示了如何使用梯度提升技术执行Bellman算子的两个步骤:应用和投影步骤。我们提出的框架在标准域上表现得相当好,而不使用领域模型和使用更少的训练轨迹。

英文摘要:

We consider the problem of Approximate Dynamic Programming in relational domains. Inspired by the success of fitted Q-learning methods in propositional settings, we develop the first relational fitted Q-learning algorithms by representing the value function and Bellman residuals. When we fit the Q-functions, we show how the two steps of Bellman operator; application and projection steps can be performed using a gradient-boosting technique. Our proposed framework performs reasonably well on standard domains without using domain models and using fewer training trajectories.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑