Modeling variable selection method based on correlation and principal component analysis

Abstract

The invention relates to a modeling variable selection method based on correlation and principal component analysis. An information entropy theory is used for calculating correlation information coefficients among influence factors, redundant variables are removed, then a principal component analysis method is used for extracting residual variable principal components, the number of modeling variables is reduced, and the method comprises the specific steps that 1, the information entropy theory is used for calculating the correlation information coefficients between to-be-selected variables and target variables; step 2, rejecting to-be-selected variables with relatively small associated information coefficients; 3, extracting residual variable principal components to be selected by utilizing a principal component analysis method; and step 4, obtaining a final modeling variable. According to the method, the number of modeling variables can be reduced on the premise that modeling variable information is reserved as much as possible, and help is provided for establishing an object model.

Description

technical field [0001] The invention relates to the technical field of data-driven modeling, in particular to a modeling variable selection method based on correlation and principal component analysis. Background technique [0002] In recent years, with the continuous development of computer and database technology, data-driven modeling technology has attracted more and more attention. However, due to the complexity of the object, there are often a large number of candidate variables that may be related to the target variable. Among them, there are correlations and couplings among some of the candidate variables, and some of the candidate variables cannot reflect the system output, and there are noises and irrelevance. or redundancy. If all possible variables to be selected are included in the model, it will greatly increase the modeling time and the uncertainty of the model, weaken the generalization ability of the model, and reduce the accuracy of the model. Therefore, in...

AI Technical Summary

Problems solved by technology

However, due to the complexity of the object, there are often a large number of candidate variables that may be related to the target variable, and there are correlations and couplings between some of the candidate variables, and some of the candidate variables cannot reflect the system output, and there are noises and irrelevance. or redundant

Images