A New Optimized Approach to Resolve a Combinatorial Problem: CoronaVirus Optimization Algorithm and Self-organizing Maps

El Majdoubi, O.; Abdoun, F.; Abdoun, O.

The optimization provides resolutions to complex combinatorial problems that generally deal within large data size and expensive operating processes. Metaheuristics target the promoting of new applied algorithms to resolve NP-Hard problems in order to improve resolutions requiring enhanced search strategies. Travelling Salesman Problem (TSP) is a common combinatorial problem, applied on several benchmark networks, namely the transportation networks and the routing vehicle problem in order to establish new intelligent computing methods as well as, to prove studies on their performances and efficiencies. Collective intelligence has proven satisfactory resolutions wherewith metaheuristics, although their algorithms complexity. The aim of this work is to solve the Euclidean TSP, classified as a NP-hard problem by means of Self-Organizing Maps (SOM) which is a Kohonen-type network. The resolution is also computed corresponding to a new bio-inspired evolutionary strategy so-called coronavirus optimization algorithm combined with SOM algorithm. The present approach is combining an unsupervised learning strategy within the new coronavirus optimization algorithm to replicate iteratively new infected individuals and to generate diversification on the search space. This new hybrid method presents a good approximative resolutions, proved by applying tests for TSPLIB instances wherein the exact optimum is defined corresponding to each TSP data. However, the present resolutions are complex specifically for large scale by means of increasing the size of input data or size parameters. © 2021, The Author(s), under exclusive license to Springer Nature Switzerland AG.