Classifier generation method and device, storage medium and electronic equipment

Abstract

The invention discloses a classifier generation method and device, a storage medium and electronic equipment, and belongs to the technical field of computers. The classifier generation method comprises the steps: acquiring an nth integrated classifier and an nth data subset, processing the nth data subset through the nth integrated classifier to obtain an nth meta-state parameter, processing the nth meta-state parameter through a meta-sampler to obtain an nth sampling weight, generating an (n+1)th data subset based on the nth sampling weight, combining the trained (n+1)th base classifier withthe nth integrated classifier to obtain an (n+1)th integrated classifier, and when n+1 is greater than a threshold, taking the (n+1)th integrated classifier as a target integrated classifier. Therefore, by constructing the meta-sampler, the optimal sampling strategy can be provided for the given task by automatically learning from the data, the performance of an ensemble learning model is effectively improved, and the problem that the machine learning effect on the category imbalance data is poor is solved.

Description

technical field [0001] The invention relates to the technical field of computers, in particular to a classifier generation method, device, storage medium and electronic equipment. Background technique [0002] Traditional machine learning methods usually assume that the amount of training data for different categories is the same, and do not consider the differences in the number of samples of different categories and the quality of identification. However, in the practical application of machine learning systems, the collected training data is often category-imbalanced, that is, in a data set, the number of samples of different categories varies greatly, resulting in a large difference in the quality of representation. For example, in financial fraud detection (normal bill / fraud bill), network intrusion detection (normal user connection / malicious connection), medical aided diagnosis (normal person / patient) and other tasks, the number of positive and negative samples is very...

AI Technical Summary

Problems solved by technology

[0003] The embodiment of the present application provides a classifier generation method, device, storage medium and electronic equipment. By constructing a parameterized, self-adaptive and learnable meta-sampler, the performance of the integrated learning model can be effectively improved and the class imbalanced data can be solved. The problem of poor machine learning performance on

Images