Abstract :

Predicting Software Defects Is Important For Making Software More Reliable Because It Finds Broken Modules Before They Are Deployed. This Study Shows A Hybrid Swarm Optimized Machine Learning Software Defect Prediction System That Combines Strong Classifiers With Metaheuristic Feature Selection. Artificial Bee Colony, Grasshopper Optimization Algorithm, GFGOA_ABC, And LGFGOA_ABC Are Some Of The Discriminative Measures Used In Feature Optimization. The Synthetic Minority Over Sampling Technique Is Used To Deal With Class Mismatch. XGBoost, Multi Layer Perceptron, And A Stacking Classifier That Combines Random Forest, XGBoost, And Decision Tree Models Are Used For Classification. The GFGOA ABC Extension Model Works Best On NASA MDP Datasets CM1, KC1, KC4, PC3, And PC5. It Gets 93.6 Percent Accuracy On CM1, 80.9 Percent Accuracy On KC1, 86.1 Percent Accuracy On KC4, 93.7 Percent Accuracy On PC3, And 84.0 Percent Accuracy On PC5. Explainable AI Methods Like LIME And SHAP Make Things Clear By Figuring Out How Features Affect Each Other. For Real-world Use, The Optimized Model Is Put Into Action Using A Flask-based Web Application That Tells Developers In Real Time Which Software Modules Are Broken And Which Ones Are Not. This Helps Them Make Reliable Quality Assurance Decisions For Large Development Teams And Continuous Integration Environments Around The World.

Published:

08-4-2026

Issue:

Vol. 26 No. 4 (2026)

Page Nos:

1680-1696

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite