Single-class classification method and classifier based on fuzzy reasoning

Abstract

The invention provides a fuzzy reasoning-based single-class classification method and classifier, which fuzzifies a feature vector, generates a fuzzy rule set and corrects the fuzzy rule set. The specific method is as follows: after a target rule set is generated, rule correction needs to be carried out if the same judgment condition appears but the judgment results are different. And the single-class classifier can accurately identify abnormal samples. Firstly, data processing is carried out, then fuzzification is carried out on the data, then a fuzzy rule set is established, and rule correction is carried out. After the rule set is established, a test sample can be used for testing the rule set, and when the rule generated after the sample of the test set is subjected to data processingof the same mapping is different from the existing rule, the sample is considered as an abnormal sample and then classified into an unknown category. According to the method, the number of data set samples in the aspect of single-class classification is increased. Meanwhile, a better classification effect can be achieved by applying a more optimized algorithm, and the training speed is increased.

Description

technical field [0001] The invention relates to the design and realization of a single-class classifier, in particular, the fuzzy reasoning in the fuzzy mathematics theory is used to realize the single-class classifier, which can be used in the field of pattern recognition. [0002] For example, the current mainstream detection method for wood species identification is the non-destructive detection method, and the present invention uses a fuzzy reasoning classifier to perform single-class classification for visible light / near-infrared spectral data of wood samples. The present invention also carries out single-class classification for the iris flower data set (Iris) and automobile data set (Car) of UCI; simultaneously also carries out single-class classification for the adult income data set (Adult) and red wine data set (Wine) of UCI . Background technique [0003] The data faced by single-class classification is complex and changeable, and there is a great possibility of ...

AI Technical Summary

Problems solved by technology

[0004] At present, there are deficiencies in the general single-class classification technology. The first deficiency is in the training set. Some of the current mainstream single-class classifiers need a large-scale training set to achieve better classification results, such as neural networks (BP Neural Networks). Network and Convolutional Neural Network, etc.), Naive Bayes, Random Forest and other algorithms, these algorithms have poor recognition results on small training sets

The second shortcoming is that in terms of training speed, many single-class classifiers require a long training time in the case of a large number of training set samples, such as neural network, support vector machine, correlation vector machine and other algorithms

Images