Software defect prediction data processing method and device and storage medium

Abstract

The invention discloses a software defect prediction data processing method. According to the method, an independent feature distribution model is established for each feature of a defect sample; andpart of the features are replaced by adopting a random variation mode so as to obtain a new defect sample, and new samples are supplemented continuously until the ratio of a non-defect sample to thedefect sample is balanced so as to obtain a processed software defect prediction data set for subsequent model training. The invention further provides a software defect prediction data processing device based on the method and a machine storage medium, the problem that in the prior art, due to the fact that the number of defect samples is smaller than that of non-defect samples, the defect samplerecognition capacity is insufficient is solved, and the software defect prediction precision is effectively improved.

Description

technical field [0001] The invention relates to a supplementary data set generation method and device, in particular to a software defect prediction data processing method, device and storage medium. Background technique [0002] Software defect prediction can help developers locate defect-prone modules in the project before the software product enters the testing stage, allocate limited testing resources more reasonably, and improve the quality of software products. In the process of software defect prediction, historical defect data is usually used to train binary classifiers to classify software modules to be predicted into defect and non-defect categories, and the classification results are used as the basis for judging the defect tendency of the module. However, in the software defect prediction data set, the number of defect samples is often far less than the number of non-defect samples, so the generated model tends to be biased towards a large number of non-defect cl...

AI Technical Summary

Problems solved by technology

This type of method can generate samples that are highly similar to the original samples and retain the original data features to the greatest extent, but because only local sample information is considered during linear interpolation, and the features are mutually restricted (since new samples can only exist in two On the connection between two parent samples, once a feature is determined, all other features cannot be changed), the generated new sample is too similar to the original sample, so the processed data set cannot effectively improve the model's identification of different defect samples. ability

Images