A Brief Analysis of Soft Computing Techniques in Software Fault Prediction

Software testing and fault prediction is major part of any software development. It not only reduces the overall cost and efforts but also increase the robustness of the system. Tester has to find or predicts the fault in advance to design premium fault free software. Soft-computing is basically is used for identifying defect in advance through techniques like Fuzzy logic, Neural Network and other metaheuristic techniques. In this paper we are doing a brief analysis of different soft computing techniques used in software fault prediction. Different datasets and parameter selection is used in the calculation of faulty module, model's efficiency and accuracy. Recent studies with their benefit and loss are being compared to find out best possibilities and combination of model for better stabilized results