Data dimension reduction method based on maximum ratio and linear discriminant analysis

Abstract

The invention relates to a data dimension reduction method based on a maximum ratio and linear discriminant analysis, and belongs to the field of image classification and pattern recognition. The method comprises the steps of constructing a data matrix, a label vector and a label matrix; calculating an intra-class covariance matrix and an inter-class covariance matrix; constructing an optimization problem of linear discriminant analysis based on the maximum ratio sum; and solving a projection matrix capable of maximizing the objective function by using an alternating optimization iterative algorithm. According to the method, the objective function of the linear discriminant analysis method based on the maximum ratio sum is established, the problem that traditional linear discriminant analysis tends to select features with small variance and weak discriminant ability is avoided, and features more beneficial to classification can be selected. The method does not depend on calculation of an inverse matrix of an intra-class covariance matrix, does not need data preprocessing, and improves the adaptability of the data dimension reduction method to original data features.

Description

technical field [0001] The invention belongs to the field of image classification and pattern recognition, in particular to a data dimensionality reduction method based on maximization ratio and linear discriminant analysis. Background technique [0002] Data dimensionality reduction technology is an important research topic in the field of image classification and pattern recognition. In the context of big data, the amount of raw data directly obtained in actual application scenarios is huge. The high dimensionality and high redundancy of these data have caused great difficulties for data storage and data processing, and have increased the demand for data storage and Handle hardware platform requirements. Data dimensionality reduction is to reduce the dimensionality of the original high-dimensional data. While ensuring that the dimensionality-reduced data still retains most of the information contained in the original data, the dimensionality of the data is reduced as much...

AI Technical Summary

Problems solved by technology

[0008] Aiming at the problem that the traditional linear discriminant analysis that has been proposed tends to select features with small variance and low discriminant power, and the solution of the optimization problem depends on the reversible intra-class covariance matrix, the present invention proposes a method based on maximizing the ratio and linear discriminant The data dimensionality reduction method of analysis; Due to the problem of low efficiency and low accuracy in the image classification method due to the imperfection of the dimensionality reduction method, the present invention proposes a method for object classification of hyperspectral images

Images