arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2407.16807cs.LGcs.AI
收藏

多目标强化学习中架构与损失函数的探索

In Search for Architectures and Loss Functions in Multi-Objective Reinforcement Learning

EPFL(洛桑联邦理工学院);CLAIRE

Mikhail Terekhov, Caglar Gulcehre

AI总结:

针对多目标强化学习(MORL)中深度函数逼近器导致学习动态不稳定的问题,本文实证探索无模型策略学习损失函数与架构选择,提出MOPPO与MOA2C方法,实验表明MOPPO能有效捕捉Pareto前沿且在多项指标上优于PCN等方法。

摘要:

多目标强化学习(MORL)对于解决现实RL问题的复杂性至关重要,这些问题通常需要在多个效用函数之间进行权衡。然而,由于基于深度学习的函数逼近器存在不稳定的学习动态,MORL极具挑战性。最常采用的研究路径是探索不同的基于价值的损失函数来解决此问题。我们的工作实证探索了无模型策略学习损失函数及不同架构选择的影响。我们引入了两种不同方法:将PPO扩展至MORL的多目标近端策略优化(MOPPO),以及在消融实验中作为简单基线的多目标优势演员评论家(MOA2C)。我们提出的方法实现简单,仅需在函数逼近器层面进行微小修改。我们在MORL的Deep Sea Treasure、Minecart和Reacher环境中进行了全面评估,表明MOPPO能有效捕捉Pareto前沿。我们广泛的消融研究和实证分析揭示了不同架构选择的影响,凸显了与Pareto条件网络(PCN)和Envelope Q-learning等流行MORL方法相比,在超体积和期望效用等MORL指标方面,MOPPO的鲁棒性与多功能性。

英文摘要:

Multi-objective reinforcement learning (MORL) is essential for addressing the intricacies of real-world RL problems, which often require trade-offs between multiple utility functions. However, MORL is challenging due to unstable learning dynamics with deep learning-based function approximators. The research path most taken has been to explore different value-based loss functions for MORL to overcome this issue. Our work empirically explores model-free policy learning loss functions and the impact of different architectural choices. We introduce two different approaches: Multi-objective Proximal Policy Optimization (MOPPO), which extends PPO to MORL, and Multi-objective Advantage Actor Critic (MOA2C), which acts as a simple baseline in our ablations. Our proposed approach is straightforward to implement, requiring only small modifications at the level of function approximator. We conduct comprehensive evaluations on the MORL Deep Sea Treasure, Minecart, and Reacher environments and show that MOPPO effectively captures the Pareto front. Our extensive ablation studies and empirical analyses reveal the impact of different architectural choices, underscoring the robustness and versatility of MOPPO compared to popular MORL approaches like Pareto Conditioned Networks (PCN) and Envelope Q-learning in terms of MORL metrics, including hypervolume and expected utility.

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。

↑