POD-MOO.m

A new Multi-Objective Optimization (MOO) algorithm with Chebyshev decomposition and Proper Orthogonal Decomposition (POD, also called Principal Components Analysis, PCA). The problem is reformulated into a set of Single-Objective Optimization (SOO) problems. New individuals are generated in the confidence ellipsoid spanned by the principle components. Magnitudes of the axes of the confidence ellipsoid are determined adaptively with an unbiased estimator. It takes around 20 iterations for the POD-MOOP algorithm to converge to the true Pareto front of the standard ZDT series test functions.
For more detailed information of the algorithm please visit the webpage https://sites.google.com/site/adloptimization/moo-with-principle-component-analysis.