IVF/NSGA-III - In Vitro Fertilization Method Coupled to NSGA-III
Sávio Menezes Sampaio, Altino Dantas, Celso G. Camilo-Junior · 2023
The In Vitro Fertilization Genetic Algorithm - IVF/GA is a promising algorithm applicable to single-objective problems, especially complex and multimodal ones. Due to the balance between its exploration and exploitation phases, and its capability to avoid local optimal, we speculate that this method can also improve Many-Objective Evolutionary Algorithms. Thus, this work proposes the adaptation of the In Vitro Fertilization method to the Many-Objective approach, with a new collection, manipulation and transfer criteria, as well as its coupling to the Many-Objective Evolutionary Algorithm NSGA-III, called IVF/NSGA-III. We evaluated the efficacy of the proposals by comparing canonic NSGA-III with memetic IVF/NSGA-III, applied to Many-Objective benchmark DTLZ. The results show that IVF/NSGA-III outperformed canonical NSGA-III in most problems addressed. The results also indicate that this proposal is promising to improve MOEAs.