A Classification Method of Portrait Data Based on Support Vector Machine

Abstract

The present invention relates to the technical field of portrait data classification, in particular to a classification method for portrait data based on a support vector machine, comprising: obtaining original portrait data, and removing noise in the original portrait data by using a CRF denoising method based on a complete random forest to obtain low Noisy portrait data set; input the low-noise portrait data set into the SVM as the training set of the support vector machine, the SVM classifies the low-noise portrait data, obtains the classified portrait data, and uses the classified portrait data as a new training set for portrait recognition . The method of the invention can effectively remove the class noise in the portrait data, solve the problem that the support vector machine SVM is difficult to classify when the training set contains more noise, improve the classification efficiency and accuracy of the portrait data, and use the classified new data set for The training set of portrait recognition indirectly improves the success rate of portrait recognition.

Description

technical field [0001] The invention relates to the technical field of portrait data classification, in particular to a support vector machine-based portrait data classification method. Background technique [0002] Support Vector Machine (SVM) is a class of generalized linear classifiers that perform binary classification on data by supervised learning. The idea of ​​support vector machine is mainly applied to the methods of solving multi-class problems, mainly including one class of residual classes, paired classification and decision tree methods, etc., and it has been mainly used in pattern recognition, such as the recognition of handwritten digits, speech recognition, etc. , and later applied to various fields such as bioinformatics, face detection, and text classification networks, and achieved good results. [0003] Face recognition technology is a technology that uses a computer to analyze a face image, extracts effective identification information from it, and uses...

AI Technical Summary

Problems solved by technology

[0005] 1. It is difficult to implement large-scale training samples: the space consumption of SVM is mainly to store training samples and kernel matrices, because SVM uses quadratic programming to solve support vectors, and solving quadratic programming will involve the calculation of m-order matrices (m is The number of samples), when the number of m is large, the storage and calculation of the matrix will consume a lot of machine memory and computing time

If the amount of data is large, the training time of SVM will be longer

[0006] 2. Difficulties in solving multi-category problems: the classic support vector machine algorithm only gives the algorithm of two-class classification, but in practical applications, it is generally necessary to solve multi-class classification problems

Images