Bi-Directional Self-Organization Technique for Enhancing the Genetic Algorithm
Kalluri sai Dinesh, R. Subramanian, Ibrahim Dweib, M. Nandhini, Mohamed Yasin Noor Mohamed, R. Rajakumar · 2018
Genetic Algorithm (GA) is well-known optimization algorithm for solving various kinds of the optimization problem. GA is based on evolutionary principles and effectively solves the large-scale problem. In addition, it incorporates the variety of hybrid techniques to achieve the best performance in complex problems. However, self-organization is one of the popular model, which acquire global order from the local interaction among the individuals. The combined version of self-organization and genetic algorithm are adopted to improve the performance in attaining the convergence. This paper proposes a bi-directional self-organization process for improving the genetic algorithm which achieves the convergence and well-balanced diversity in the population. The experimentation is conducted on the standard test-bed of travelling salesman problem and instances are obtained from TSPLIB. Thus, the proposed algorithm has shown its dominance with the existing classical GA in terms of various parameter metrics.