Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
4963166 | Applied Soft Computing | 2017 | 41 Pages |
Abstract
A surrogate-assisted (SA) evolutionary algorithm for Multiobjective Optimization Problems (MOOPs) is presented as a contribution to Soft Computing (SC) in Artificial Intelligence (AI). Such algorithm is grounded on the cooperation between a “pure” evolutionary algorithm and a Kriging based algorithm featuring the Expected Hyper-Volume Improvement (EHVI) metric. Comparison with state-of-art pure and Kriging-assisted algorithms over two- and three-objective test functions have demonstrated that the proposed algorithm can achieve high performance in the approximation of the Pareto-optimal front mitigating the drawbacks of its parent algorithms.
Keywords
Related Topics
Physical Sciences and Engineering
Computer Science
Computer Science Applications
Authors
Giovanni Venturelli, Ernesto Benini, Åukasz Åaniewski-WoÅÅk,