An incremental software defect detection model based on support vector machine

Abstract

Purpose

(1) developing a dynamic and progressive software defect prediction model to successfully manage novel and huge amounts of software defect data and lessen the computational time. (2) to avoid the great diminish of static batch learning algorithms efficiency once the amount of data achieves a certain level.

Design/methodology/approach

This study explores the proficiency of the incremental classification based approach to elaborate anincremental software defect prediction system which helps recognizing and treating real-time software data streams.

Findings

The proposed method, as demonstrated by experimental results, is clearly competitive with the relevant two-class classifiers currently in use for software defect diagnosis. Detailed experimental findings clearly demonstrated the performance and efficiency of the suggested software defect detection approach: Incremental Discriminant-based Support Vector Machine (IDSVM) to differentiate between defective and non-defective objects.

Originality/value

To the best of our knowledge, this is the first a real-time prediction method that investigates incremental classification in software defect prediction research

Citation

Zaibi, D., Salhi, M., Tbarki, K. and Ksantini, R. (2024), "An incremental software defect detection model based on support vector machine", Engineering Computations, Vol. ahead-of-print No. ahead-of-print. https://doi.org/10.1108/EC-11-2023-0799