Improving Local Optima of Local Search with Adjustable Multipliers
Zongmei Zhang, Hiroki Tamura, Zheng Tang, Jun Ma
Abstract
Zongmei Zhang, Hiroki Tamura, Zheng Tang, Jun Ma
Abstract
In this paper, we give two improved algorithms of Guided Local Search (GLS) to improve the local optima of local search. In the GLS-like algorithm, a new penalty principle is proposed to further improve the effectiveness of GLS. The Objective function Adjustment (OA) algorithm is an improved algorithm of GLS-like using multipliers which can be adjusted during the search process. The simulation results based on some TSPLIB benchmark problems showed that the OA algorithm could find better solutions than the local search, guided local search, Tabu Search and GLS-like.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
In this paper, we give two improved algorithms of Guided Local Search (GLS) to improve the local optima of local search. In the GLS-like algorithm, a new penalty principle is proposed to further improve the effectiveness of GLS. The Objective function Adjustment (OA) algorithm is an improved algorithm of GLS-like using multipliers which can be adjusted during the search process. The simulation results based on some TSPLIB benchmark problems showed that the OA algorithm could find better solutions than the local search, guided local search, Tabu Search and GLS-like.
Key concepts: Guided Local Search, Local optimum, Tabu search, Local search (optimization), Hill climbing, Mathematical optimization, Benchmark (surveying), Best-first search