The Pareto Differential Evolution Algorithm

Abstract

The use of evolutionary algorithms (EAs) to solve problems with multiple objectives (known as Vector Optimization Problems (VOPs)) has attracted much attention recently. Being population based approaches, EAs offer a means to find a group of pareto-optimal solutions in a single run. Differential Evolution (DE) is an EA that was developed to handle optimization problems over continuous domains. The objective of this paper is to introduce a novel Pareto Differential Evolution (PDE) algorithm to solve VOPs. The solutions provided by the proposed algorithm for five standard test problems, is competitive to nine known evolutionary multiobjective algorithms for solving VOPs.

Automatic Detection of Semantic Primitives with Multi-Objective Bioinspired Algorithms and Weighting Algorithms

“…Due to the characteristics of the problem to be solved, a multi-objective ED algorithm was implemented, specifically, the Pareto Differential Evolution (PDE) [1], which is a modification of the original ED and whose algorithm is presented below. The next considerations were applied to the multi-objective algorithm:…”

Benchmarking a Wide Spectrum of Metaheuristic Techniques for the Radio Network Design Problem

“…It is currently used in a wide range of optimization problems, including multiobjective optimization [80]. Generally, the function to be optimized F is computed by means of optimizing the values of its parameters, where X denotes a vector composed of n param objective function parameters.…”

Automatic clustering with multi-objective Differential Evolution algorithms

“…We compared the performances of four recently developed and popular MO-variants of DE: the Pareto DE (PDE) [3], the Multi-objective DE (MODE) [4], DE for Multi-objective Optimization (DEMO) [5], and NonDominated Sorting DE (NSDE) [6]. Due to space limitations, we briefly discuss here the outline of these algorithms instead of reiterating through their details available in cited literatures.…”

“…This paper primarily compares the performances of four most representative multi-objective DE algorithms on the multi-objective fuzzy clustering problem. The multi-objective DE-variants considered here are namely the Pareto DE (PDE) [5], the Multi-objective DE (MODE) [6], DE for Multi-objective Optimization (DEMO) [7], and Non-Dominated Sorting DE (NSDE) [8]. Since DE, by nature, is a real-coded population-based optimization algorithm, we here resort to centroid-based representation scheme for the search variables.…”