KAN在离线强化学习中表现接近MLP，参数量却大幅减少。
KAN v.s. MLP for Offline Reinforcement Learning
- 用KAN替代传统MLP作为强化学习的主干网络
- 在D4RL基准上达到与MLP相当性能，参数量减少近一半
- 适合对模型轻量化和可解释性有要求的任务
Kolmogorov-Arnold Networks（KAN）是机器学习领域新兴的神经网络架构，研究者关注其是否能成为广泛使用的多层感知机（MLP）的有力替代。多个领域的实验表明，基于KAN的机器学习方法在性能上可媲美甚至超越基于MLP的方法，同时参数规模更小且更具可解释性。本文探索将KAN融入离线强化学习中的策略网络（actor）和价值网络（critic），在经典的D4RL基准上评估了多种基于KAN与MLP的保守Q学习（CQL）方法的性能、参数规模和训练效率。结果表明，KAN可在保持接近MLP性能的同时，显著减少参数量。这为根据离线强化学习任务需求选择基础网络提供了新选项。
原文摘要 · Abstract (English)
Kolmogorov-Arnold Networks (KAN) is an emerging neural network architecture in machine learning. It has greatly interested the research community about whether KAN can be a promising alternative of the commonly used Multi-Layer Perceptions (MLP). Experiments in various fields demonstrated that KAN-based machine learning can achieve comparable if not better performance than MLP-based methods, but with much smaller parameter scales and are more explainable. In this paper, we explore the incorporation of KAN into the actor and critic networks for offline reinforcement learning (RL). We evaluated the performance, parameter scales, and training efficiency of various KAN and MLP based conservative Q-learning (CQL) on the the classical D4RL benchmark for offline RL. Our study demonstrates that KAN can achieve performance close to the commonly used MLP with significantly fewer parameters. This provides us an option to choose the base networks according to the requirements of the offline RL tasks.
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。