Beyond Hyper-Heuristics: A Squared Hyper-Heuristic Model for Solving Job Shop Scheduling Problems

Hyper-heuristics (HHs) stand as a relatively recent approach to solving optimization problems. There are different kinds of HHs. One them deals with how low-level heuristics must be combined deliver an improved solution set problem instances. Literature commonly refers selection hyper-heuristics . their advantages is that the strengths each heuristic can fused into high-level solver. However, one drawbacks sometimes this generalization scheme does not suffice. Additionally, it easy reuse these HHs since model cannot easily tweaked. So, in work, we develop hyper-heuristic additional layer generalization. The rationale behind preserve general structure selecting adequate solver for particular situation but use instead heuristics. We call Squared Hyper-Heuristic (SHH). To validate our proposal, pursue four-stage methodology covers several testing scenarios. Our data reveal that, under proper conditions, outperform base Moreover, flexible enough allow increased number layers so complexity final tuned. instances used train stage model, thus setting groundwork developing transfer learning hyper-heuristics.