Novel performance metrics for robust multi-objective optimization algorithms

Seyedali Mirjalili, Andrew Lewis

Research output: Contribution to journalArticlepeer-review

63 Citations (Scopus)


Performance metrics are essential for quantifying the performance of optimization algorithms in the field of evolutionary multi-objective optimization. Such metrics allow researchers to compare different algorithms quantitatively. In the field of robust multi-objective optimization, however, there is currently no performance metric despite its significant importance. This motivates our proposal of three novel specific metrics for measuring the convergence, coverage, and success rate of robust Pareto optimal solutions obtained by robust multi-objective algorithms. The proposed metrics are employed to quantitatively evaluate and compare Robust Multi-objective Particle Swarm Optimization (RMOPSO) and Robust Non-dominated Sorting Genetic Algorithm (RNSGA-II) on seven selected benchmark problems. The results show that the proposed metrics are effective in quantifying the performance of robust multi-objective algorithms in terms of convergence, coverage, and the ratio of the robust/non-robust Pareto optimal solutions obtained.

Original languageEnglish
Pages (from-to)1-23
Number of pages23
JournalSwarm and Evolutionary Computation
Publication statusPublished - 1 Apr 2015
Externally publishedYes


  • Robust multi-objective optimization Multi-objective optimization Performance metric Convergence metric Coverage metric Success ratio


Dive into the research topics of 'Novel performance metrics for robust multi-objective optimization algorithms'. Together they form a unique fingerprint.

Cite this