Correlation Analysis of Stocks and PMI Index Based on Logistic Regression Model

Abstract

In order to explore the correlation between stocks and the PMI index, based on the generalized logistic loss and margin distribution, this paper designs a margin distribution logistic regression model that is easy to optimize, has robustness, and generalization ability, and gives a multiclass margin distribution logistic regression framework. This framework can be used to perform two-classification, multiclassification, and feature selection tasks. Moreover, this paper gives a training algorithm for margin distribution logistic regression on large-scale data sets through the pairwise stochastic gradient descent method. In addition, this paper combines the logistic regression model to construct a correlation analysis model between stocks and PMI index and uses the PMI data of the National Bureau of Statistics as a sample to design experiments to verify the performance of the system model constructed in this paper. From the experimental analysis, it can be seen that the algorithm constructed in this paper has a certain effect, and the strong correlation between PMI and stocks has been further verified.