A Dynamic Resource Allocation Strategy with Reinforcement Learning for Multimodal Multi-objective Optimization

作     者：Qian-Long Dang Wei Xu Yang-Fei Yuan Qian-Long Dang;Wei Xu;Yang-Fei Yuan

作者机构：School of Mathematics and StatisticsXidian UniversityXi’an 710126China 

出 版 物：《Machine Intelligence Research》 (机器智能研究（英文版）)

年 卷 期：2022年第19卷第2期

页      面：138-152页

核心收录：

学科分类：12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 081104[工学-模式识别与智能系统] 08[工学] 0835[工学-软件工程] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

主　　题：Multimodal multi-objective optimization(MMO) dynamic resource allocating strategy(DRAS) reinforcement learning(RL) decision space partition zoning search 

摘      要：Many isolation approaches, such as zoning search, have been proposed to preserve the diversity in the decision space of multimodal multi-objective optimization(MMO). However, these approaches allocate the same computing resources for subspaces with different difficulties and evolution states. In order to solve this issue, this paper proposes a dynamic resource allocation strategy(DRAS)with reinforcement learning for multimodal multi-objective optimization problems(MMOPs). In DRAS, relative contribution and improvement are utilized to define the aptitude of subspaces, which can capture the potentials of subspaces accurately. Moreover, the reinforcement learning method is used to dynamically allocate computing resources for each subspace. In addition, the proposed DRAS is applied to zoning searches. Experimental results demonstrate that DRAS can effectively assist zoning search in finding more and better distributed equivalent Pareto optimal solutions in the decision space.