ENSEMBLE LEARNING-BASED DETECTION OF HIGH-RISK SOFTWARE DEFECTS IN ENTERPRISE APPLICATIONS

Abstract

Enterprise applications are increasingly vulnerable to software defects due to growing system complexity, rapid development cycles, cloud adoption, and extensive dependency usage. High-risk software defects can introduce security weaknesses, operational failures, and reliability issues that negatively impact business continuity. Traditional defect detection approaches based on static analysis, rule-based techniques, and manual inspection often face limitations in identifying complex defect patterns and prioritizing critical issues. This paper proposes an Ensemble Learning-Based Detection Framework for HighRisk Software Defects in Enterprise Applications by integrating ensemble machine learning, software metrics analysis, vulnerability intelligence, and automated software testing. The proposed framework combines multiple predictive models to analyze source code characteristics, historical defect data, testing outcomes, dependency information, and software quality indicators. Ensemble learning improves defect classification accuracy by leveraging the strengths of multiple algorithms. The framework identifies high-risk defects, prioritizes remediation efforts, and supports continuous software quality improvement. Experimental evaluation demonstrates improvements in defect prediction accuracy, risk classification, falsepositive reduction, and enterprise software reliability.