Software Defect Prediction Method Based on Neighborhood Embedding Protection Algorithm Support Vector Machine

Abstract

The invention proposes a software defect prediction method based on a neighborhood embedding protection algorithm support vector machine, which is used to solve the problem of software measurement data redundancy. Including: select training set X from software defect prediction data set 1 and the test set X 2 ; Using the NPE algorithm for the training set X 1 and the test set X 2 Carry out dimensionality reduction; the training set Y after dimensionality reduction 1 As the training input set, use the support vector machine SVM for training to obtain the trained defect prediction model; the dimensionality-reduced test set Y 2 As a test input set, use the trained defect prediction model to make predictions, and compare the prediction results with the actual results. If the prediction results meet the termination conditions, the software defect prediction model at this time is the optimal software defect prediction model; otherwise, restart Perform SVM training for optimization.

Description

technical field [0001] The invention relates to a software defect prediction method based on a neighborhood embedding protection algorithm support vector machine, belonging to the technical field of software prediction. Background technique [0002] Software defects: IEEE729-1983 has a standard definition of defects. From the internal point of view of the product, defects are various problems such as errors and defects in the process of software product development or maintenance; from the external point of view of the product, defects are the software system and its original The failure or violation of a certain function that needs to be realized. [0003] Static prediction: Static software defect prediction technology is an earlier prediction technology that is currently researched and applied most. It is mainly based on extracting software-related measurement information to establish a corresponding prediction model for defect prediction. [0004] Prediction model based ...

AI Technical Summary

Problems solved by technology

[0012] The present invention proposes a software defect prediction method based on the support vector machine based on the neighborhood embedding protection algorithm, which is used to solve the problem of software measurement data redundancy, so that the final defect distribution prediction result has a higher accuracy than other traditional prediction technologies Accuracy

Images