On the performance metrics of multiobjective optimization

Authors:
Shi Cheng;Yuhui Shi;Quande Qin
Affiliations:
Department of Electrical Engineering and Electronics, University of Liverpool, Liverpool, UK,Department of Electrical & Electronic Engineering, Xi'an Jiaotong-Liverpool University, Suzhou, Chi ...;Department of Electrical & Electronic Engineering, Xi'an Jiaotong-Liverpool University, Suzhou, China;College of Management, Shenzhen University, Shenzhen, China
Venue:
ICSI'12 Proceedings of the Third international conference on Advances in Swarm Intelligence - Volume Part I
Year:
2012

Citing 10
Cited 0

Multiobjective optimization problems with complicated Pareto sets, MOEA/D and NSGA-II

IEEE Transactions on Evolutionary Computation
Approximating the set of Pareto-optimal solutions in both the decision and objective spaces by an estimation of distribution algorithm

IEEE Transactions on Evolutionary Computation
A favorable weight-based evolutionary algorithm for multiple criteria problems

IEEE Transactions on Evolutionary Computation
A fast and elitist multiobjective genetic algorithm: NSGA-II

IEEE Transactions on Evolutionary Computation
Properties of an adaptive archiving algorithm for storing nondominated vectors

IEEE Transactions on Evolutionary Computation
Performance assessment of multiobjective optimizers: an analysis and review

IEEE Transactions on Evolutionary Computation
The balance between proximity and diversity in multiobjective evolutionary algorithms

IEEE Transactions on Evolutionary Computation
A faster algorithm for calculating hypervolume

IEEE Transactions on Evolutionary Computation
Diversity Management in Evolutionary Many-Objective Optimization

IEEE Transactions on Evolutionary Computation
A Fast Way of Calculating Exact Hypervolumes

IEEE Transactions on Evolutionary Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

Multiobjective Optimization (MOO) refers to optimization problems that involve two or more objectives. Unlike in the single objective optimization, a set of solutions representing the tradeoff among the different objects rather than an unique optimal solution is sought in MOO. How to measure the goodness of solutions and the performance of algorithms is important in MOO. In this paper, we first review the performance metrics of multiobjective optimization and then classify variants of performance metrics into three categories: set based metrics, reference point based metrics, and the true Pareto front/set based metrics. The properties and drawbacks of different metrics are discussed and analyzed. From the analysis of different metrics, an algorithm's properties can be revealed and more effective algorithms can be designed to solve MOO problems.