A new vision of social behavior on genetic algorithm performance
Andreea Tatar, Nicolae Fat, Adrian Petrovan, Oliviu Matei

Abstract. Genetic algorithms (GAs) are typically described in terms of mutation, crossover and selection, but the social dynamics that arise inside the population are often abstracted away. This paper proposes a new vision in which the performance of genetic algorithms is analysed through the lens of social behavior of individuals. Individuals are characterised by behavioural traits that influence how they interact during crossover, how information is exchanged and how solutions diversify. Experimental results on classical combinatorial optimisation problems indicate that incorporating social-behavior inspired operators leads to improved exploration capabilities and to better-quality solutions compared with traditional GA implementations.

Keywords: genetic algorithms; social behavior; metaheuristics; population dynamics; combinatorial optimisation

Andreea Tatar, Nicolae Fat, Adrian Petrovan, Oliviu Matei, "A new vision of social behavior on genetic algorithm performance", Proc. 19th SOCO Int. Conf. on Soft Computing Models in Industrial and Environmental Applications, Springer, 2024, 2023.

Reference: Proc. 19th SOCO Int. Conf. on Soft Computing Models in Industrial and Environmental Applications, Springer, 2024.

Other publications