Unbalanced data distribution-based multi-heterogeneous base classifier fusion classification method

Abstract

The invention discloses an unbalanced data distribution-based multi-heterogeneous base classifier fusion classification method, and relates to an unbalanced data classification technology in the field of data mining. The method comprises the following steps of: preprocessing a sample by using a difference sampling rate-based resampling algorithm, including an oversampling and an under-sampling process, thereby distributing different samples to be classified for different base classifiers; calculating a classification error rate of each base classifier and further calculating the corresponding weight; counting respective results by an oversampling expert and an under-sampling expert; and fusing the final prediction result according to a classification strategy function to obtain the category of the sample. By using the multi-heterogeneous base classifier fusion classification method, important characteristics of a few types of samples are found in mass data, and the accuracy of the few types of samples can be effectively improved, so that the aim of improving the integral classification accuracy of a data set is fulfilled.

Description

technical field [0001] The invention relates to the technical field of data mining, in particular to a multi-heterogeneous base classifier fusion classification method based on unbalanced data distribution. Background technique [0002] In recent years, with the deepening of data mining research and the continuous expansion of applications, more and more researchers have increasingly felt that in some complex data environments, traditional data mining technology has been difficult to adapt to the constant changes in the new situation. Among them, the data mining method for unbalanced data distribution environment has gradually become a hot issue in this field. Since Nathalie Japkowicz comprehensively proposed learning for unbalanced data sets, the problem of data classification based on unbalanced data distribution environments has become a dedicated research topic and one of the focus of future research for the first time. The traditional classification method focuses on i...

AI Technical Summary

Problems solved by technology

However, these methods are only suitable for normal data sets, and are not suitable for unbalanced data sets. According to the current research progress, base classifier fusion methods suitable for unbalanced data distribution environments are still rare, especially in the algorithm Diversity and classification accuracy, especially in improving the accuracy of minority class samples, still face the bottleneck problem that is difficult to break through

Images