Principal Component Analysis Based on Quantum Genetic Algorithm with T-Distribution Parameters

Machine learning is a important part of the field of artificial intelligence. Principal component analysis (PCA) is a classical linear dimensionality reduction algorithm in the field of machine learning, which is usually solved by gradient ascending method. As the traditional gradient ascending method requires the convexity of the objective function, it is easy to be trapped into the local maximum value. Therefore, in order to solve the limitation of being trapped into local optimal solution, it is meaningful to study the performance optimization for principal component analysis. In this paper, a novel hybrid algorithm which employs quantum genetic algorithm (QGA) with t-distribution parameters in traditional PCA has been proposed. Via simulation analysis, it has been verified that the proposed hybrid algorithm has better performance than the single PCA algorithm.