Unified Approach to Software Reliability through Predictive Testing and Fault Tolerance Models

Strengthening Software Reliability: Predictive Testing and Fault Tolerance Models

Abstract

A series of interesting collaborative research projects led by Assoc. Prof. Dr. Siti Hafizah Ab Hamid is making significant strides in improving software reliability. These PhD and Master research works address key challenges such as defect prediction, fault tolerance, and ambiguity detection, creating a unified approach to enhancing software quality throughout the development lifecycle. The research begins with a focus on Just-in-Time Software Defect Prediction, utilizing Deep Q-Networks (DQN) to improve the detection of defects during the code change process. This model prioritizes changes likely to contain defects, reducing unnecessary code reviews and ensuring that critical issues are addressed efficiently. As a result, resources are allocated more effectively, boosting overall development efficiency.