Relevance vector machine-based multi-class data classifying method

Abstract

The invention provides a relevance vector machine-based multi-class data classifying method, which mainly solves the problem that the traditional multi-class data classifying method cannot integrally solve classifying face parameters and needs proximate calculation. The relevance vector machine-based multi-class data classifying method comprises a realizing process comprising the following steps of: partitioning a plurality of multi-class data sets and carrying out a normalizing pretreatment; determining a kernel function type and kernel parameters; setting basic parameters; calculating the classifying face parameters; calculating lower bounds of logarithms and solving variant values of the lower bounds of the logarithms and adding 1 to an iterative number; if the variant values of the lower bounds of the logarithms are converged or the iterative number reaches iterating times, finishing updating the classifying face parameters, and otherwise, continuing to updating; and obtaining a prediction probability matrix according to the updated classifying face parameters, wherein column numbers corresponding to a maximum value of each row of the matrix compose classifying classes for testing the data sets, and samples which have the prediction probability less than a false-alarm probability and the detection probability corresponding to a false-alarm probability value set in a curve are rejected. The relevance vector machine-based multi-class data classifying method has the advantages of obtaining classification which is comparable to that of an SVM (Support Vector Machine) by using less relevant vectors and rejecting performance and can be used for target recognition.

Description

technical field [0001] The invention belongs to the technical field of data processing and relates to data classification, in particular to a recognition and classification method for multi-type data, which is used in target recognition. Background technique [0002] Data classification is used to distinguish different target data, and to distinguish different target data as much as possible, so that each target data can be identified in a large number of different target data. The current data classification method is mainly to study the problem of two types of data classification, and the two types of data classification methods mainly include support vector machine method and correlation vector machine method. Support vector machine (SVM) was first proposed by Cortes and Vapnik in 1995, and it shows many unique advantages in solving small sample, nonlinear and high-dimensional pattern recognition. SVM was originally used to solve two-class data classification problems an...

AI Technical Summary

Problems solved by technology

Although the number of classifiers is large, the total time used to calculate the classification planes of these classifiers is less than that of the one-to-many method, but there is also classification overlap.

[0007] The quadratic programming algorithm is to combine the parameter solutions of K classification surfaces into one optimization problem during training, and solve the parameters required for the optimization problem through the quadratic programming method. The same method as the one-to-many algorithm is used in the test. Judgment method, that is, for an input sample, the classification result is the category with the largest output value of each sub-classifier. This algorithm is not suitable for classifying data with a large number of categories.

This method assumes Laplacian prior information with sparse characteristics, and constructs a classifier under the maximum a posteriori criterion, which is insufficient for approximate calculation.

Images