Heterogeneous Cross-Project Software Defect Prediction Method Based on Kernel Mapping Migration Integration

Abstract

The invention discloses a heterogeneous cross-project software defect prediction method based on nuclear spectrum mapping migration integration, which includes preprocessing source data based on an over-sampling unbalanced learning process; the source data is historical defect data of software projects; According to the difference information between the source data and the target data distribution, and the information loss caused by spectral mapping, the target function is constructed; the target data is the heterogeneous cross-project software to be predicted; the original space is mapped to multiple high-level by multi-core learning dimensional space, optimize the objective function on each high-dimensional space to find the optimal common subspace, train a classifier on each common subspace; use ensemble learning to comprehensively integrate all the classifiers , generate a prediction model; predict the label of the target data according to the prediction model. This method has a higher defect prediction accuracy and is beneficial to improve the work efficiency of software testing.

Description

technical field [0001] The invention relates to the technical field of software defect prediction, in particular to a heterogeneous cross-project software defect prediction method based on nuclear spectrum mapping migration integration. Background technique [0002] Software defect prediction refers to using the historical defect data of a given software project to find the mapping relationship between software module metrics and software defects through statistical or machine learning methods, and then predict the defect status of new modules of the project. If there is not enough historical defect data, it is necessary to utilize cross-project software defect prediction techniques. Cross-project software defect prediction refers to the use of historical defect data (source data) of other software projects to establish a prediction model, and then use it for new software projects (target data) to predict its defect status. Cross-project software defect prediction can be di...

AI Technical Summary

Problems solved by technology

[0006] (3) No consideration of covering the optimal common feature subspace through multiple subspaces

[0007] Therefore, the above problems make it difficult for the prediction performance of the existing prediction methods to achieve a more ideal prediction effect.

Images