A multi-objective hyper-heuristic algorithm based on adaptive epsilon-greedy selection

Abstract A variety of meta-heuristics have shown promising performance for solving multi-objective optimization problems (MOPs). However, existing may the best on particular MOPs, but not perform well other MOPs. To improve cross-domain ability, this paper presents a hyper-heuristic algorithm based adaptive epsilon-greedy selection (HH_EG) select and combine low-level heuristics (LLHs) during evolutionary procedure, also proposes an strategy. The proposed can solve from varied domains by simply changing LLHs without redesigning high-level Meanwhile, HH_EG does need to tune parameters, is easy be integrated with various indicators. We test classical DTLZ suite, IMOP many-objective MaF suite real-world problem. Experimental results show effectiveness in combining advantages each LLH problems.