Population-Based Redundancy Control in Genetic Algorithms: Enhancing Max-Cut Optimization

The max-cut problem is a well-known topic in combinatorial optimization, with a wide range of practical applications. Given its NP-hard nature, heuristic approaches—such as genetic algorithms, tabu search, and harmony search—have been extensively employed. Recent research has demonstrated that harmony search can outperform genetic algorithms by effectively avoiding redundant searches, a strategy similar to tabu search. In this study, we propose a modified genetic algorithm that integrates tabu search to enhance solution quality. By preventing repeated exploration of previously visited solutions, the proposed method significantly improves the efficiency of traditional genetic algorithms and achieves performance levels comparable to harmony search. The experimental results confirm that the proposed algorithm outperforms standard genetic algorithms on the max-cut problem. This work demonstrates the effectiveness of combining tabu search with genetic algorithms and offers valuable insights into the enhancement of heuristic optimization techniques. The novelty of our approach lies in integrating solution-level tabu constraints directly into the genetic algorithm’s population dynamics, enabling redundancy prevention without additional memory overhead, a strategy not previously explored in the proposed hybrids.