Method for classifying unbalanced data sets

Abstract

The invention discloses a method for classifying unbalanced data sets. The method is applied to the fields of network intrusion detection, animal age prediction, vehicle performance evaluation and thelike. For the problem that the classification precision of minority classes is low in the prior art is solved, according to the invention, on the basis of original training data, the relation betweenminority classes and majority classes in an original data set is utilized, a SMOTE and K nearest neighbor algorithm is utilized to process an original training data set to construct a new set, and the set focuses on the minority classes and the majority class samples related to the minority classes; according to the method, two random forests with the same size are constructed according to original training data and the new set, decision trees in the two forests are combined into a large forest, a test set is tested together to obtain a classification result, and the obtained classification precision is greatly improved compared with the prior art.

Description

technical field [0001] The invention belongs to the fields of network intrusion detection, animal age prediction, vehicle performance evaluation, etc., and particularly relates to a technology for classification of unbalanced data sets. Background technique [0002] Classification is one of the important research directions in the field of machine learning. After years of development, some relatively mature algorithms have been formed and have been successfully applied in practice. These traditional classification algorithms take classification accuracy as the greatest goal, and assume that the number of samples in each category in the data set is basically balanced. However, there is a situation in practical problems: a data set contains two types of data, and the number of samples contained in one type of data is far less than that of the other type of data. The former is called a minority. The latter class is called the majority class. Due to the difficulties encountere...

AI Technical Summary

Problems solved by technology

That is to say, this method only focuses on the processing of data sets, and does not integrate and improve with subsequent algorithms, which also has certain limitations.

Images