Principal Component Analysis with Quantum Whale Optimization Algorithm

Principal component analysis (PCA) is a classical supervised linear dimension reduction algorithm in machine learning. PCA solves the optimal dimension reduction direction by maximizing the variance of projection points. The classical methods to realize optimization are gradient ascend method or stochastic gradient ascend method depending on gradient information. However, the above methods are easy to fall into the trap of local extremum, which is not conducive to global optimization. In this paper, the quantum whale algorithm is introduced into the optimization part of PCA, which enhances the parallelism and globality of optimization, and gets rid of the dependence on gradient.