Principal Component Analysis Based on Artificial Fish Swarm with T-Distribution Parameters

It is often necessary to collect a large amount of data for analysis to find the regularities in many fields of research. and there is no doubt that the multi-variable large sample data set contains abundant information and plays an important role in the research process. Processing such large sample datasets can always be complex, and dimensionality reduction is always employed to build better datasets, and to improve the efficiency of subsequent research work. In this paper, principal component analysis (PCA) is used to simplify the data structure, to transform or to project the original data into a new feature space so as to reduce the dimension of the target data set. Conventional principal component analysis (PCA) uses gradient ascending method to solve the optimal value of parameter, which has some drawbacks. In this paper, artificial fish swarm, which is a swarm intelligence algorithm has been introduced in PCA to solve the optimal parameter value. The artificial fish swarm algorithm is used to search the optimal solution randomly and globally through the simple iterative optimization of each individual, which overcomes the disadvantage that the traditional gradient ascending method is easy to fall into the local extremum trap. The effectiveness of proposed algorithm in feature analysis of data dimensionality reduction has been verified via numerical experiments.