The Role of Elitism in Multiobjective Optimization with Evolutionary Algorithms

This paper studies the influence of elitism in Evolutionary Multiobjective Optimization. The true efficiency of inserting elitism in these algorithms was not yet sufficiently developed. Many algorithms for multiobjective optimization use different methods to keep the best individuals (the elite) founded on the duration of the search process. The problem is how this elitism can be preserved and, more than that, how this elitism can be efficiently incorporated in evolutionary algorithms. The paper presents some algorithms which use elitism and some algorithms which not use elitism and realizes a comparison of some algorithms for each category.