Contributions à la résolution des problèmes d’optimisation multiobjectif : Application à l’apprentissage automatique et aux problèmes socio-économique

Abstract

In an optimization context, there are different real-world problems that face many difficulties. Indeed, decision-makers sometimes target several objectives at the same time, increasing the search space as well as the complexity of the problem. Multi-objective optimization provides an answer to similar problems in order to balance the different objectives. Considering that objectives are often contradictory, solving such problems using classical methods is very complicated or even impossible, given the difficulties that this type of modeling generates.

This work aims to model and solve several multi-objective models, in particular, the portfolio management problem. This problem presents an interesting example of conflicting objectives, as we aim to maximize profit