Index Optimization Based on Gaussian Mixture Model

This article proposed a nonparametric test method by using Gaussian mixture model application idea for reference to evaluating the construction of index architecture. First use Gaussian mixture model to cluster analysis the index architecture. Second use rank correlation coefficient of nonparametric test method to select index architecture. Then determined the evaluate index architecture and guarantee the comprehensiveness and representativeness of the index architecture. The example shows that the new method is simple and practical and has high classification precision and is useful for index optimization.