Application of the algorithm based on the PSO and improved SVDD for the personal credit rating

Considering the question of personal credit rating, this paper proposes a hybrid method for credit assessment based on an improved Support Vector Data Description (SVDD) algorithm combined with the particle swarm optimization (PSO) algorithm. First, the paper carries out data preprocess, and then it solves the two problems: parameters optimization and feature selection at the same time using the PSO algorithm combined with the improved SVDD algorithm and assesses the credit data using the optimized parameters and features. Finally, the method constructed is tested through two data sets in practice, and the results show that the hybrid method constructed in this paper can obtain higher classification accuracy compared with some other existing credit scoring methods.