Abstract:In terms of generating offspring solutions, simulated binary crossover (SBX) and differential evolution (DE) are two of the most representative reproduction operators in evolutionary multi-objective algorithms (EMOAs). However, they are found as less effective on multi-objective problems (MOPs) with complicated Pareto optimal set (PS). Under mild conditions, the PS of an MOP is an <math>(m−1)</math>-dimensional piecewise continuous manifolds where <math>m</math> is the number of objectives. Inspired from this regularity property, this study proposes a simple yet effective reproduction operator, namely, PCA-assisted reproduction (PCA-ar). Specifically, the PCA-ar first applies principal component analysis (PCA) method to construct a new decision space with reduced number of dimensions based on the information extracted from several well converged solutions. The PS is then estimated by a hyperplane in the new decision space. To this end, new offspring are sampled from the estimated PS, and then re-converted to the original decision space for fitness calculation. In order to systematically examine the effectiveness of the PCA-ar operator, we integrate it into NSGA-II and MOEA/D, and compare the derived algorithms, nNSGA-II and nMOEA/D, with their original versions (NSGA-II with SBX operator and MOEA/D with DE operator) as well as the regularity model based multi-objective estimated distribution algorithm (RM-MEDA) on the modified DTLZ benchmarks with up to 8 objectives. Experimental results show that nNSGA-II and nMOEA/D outperform the competitor EMOAs for most of problems, which indicate that the PCA-ar is effective. Lastly, the PCA-ar is also demonstrated to have good scalability to the number of decision variables.

A Novel Path-Based Reproduction Operator for Multi-Objective Optimization

An New Evolutionary Multi-Objective Optimization Algorithm

An Elite Gene Guided Reproduction Operator for Many-Objective Optimization

Novel hybrid evolutionary algorithm for multi-objective optimization problem

An Evolutionary Many-Objective Algorithm Based on a Novel Tournament Selection Strategy

PCA-assisted reproduction for continuous multi-objective optimization with complicated Pareto optimal set

A Two-stage Evolutionary Framework For Multi-objective Optimization

A decomposition-based multi-objective evolutionary algorithm with Q-learning for adaptive operator selection

Regularity Evolution for Multiobjective Optimization

Multiobjective evolutionary algorithms: A survey of the state of the art

An Analysis on Recombination in Multi-Objective Evolutionary Optimization

A New Hyper-Heuristic Multi-Objective Optimisation Approach Based on MOEA/D Framework

Decomposition Multi-Objective Evolutionary Optimization: From State-of-the-Art to Future Opportunities

A Projection-Based Evolutionary Algorithm for Multi-Objective and Many-Objective Optimization

A Grey Prediction‐Based Reproduction Strategy for Many‐Objective Evolutionary Algorithm

A New Evolutionary Approach to Multiparty Multiobjective Optimization.

Non-Elitist Evolutionary Multi-Objective Optimisation: Proof-of-Principle Results

A coevolutionary algorithm using multi-operator ensemble for many-objective optimisation problems

An effective offspring generation strategy for many-objective optimization driven by knee points under variable classification

A Survey on Search Strategy of Evolutionary Multi-Objective Optimization Algorithms

Deep multi-layer perceptron-based evolutionary algorithm for dynamic multiobjective optimization