SOFTWARE DEFECT PREDICTION: PAST PRESENT AND FUTURE

Software development calls for several defect prediction methodologies using critical parameters such as review effort measurement, test effort estimation, phase gate containment, change request cost, re-usability, size and quality to improve the quality of deliverables. Nonetheless, a lot of these methodologies are actually in development stages and further research is required to produce a strong and dependable model. Many research centers have started more research projects in these research areas. Through this study, we investigated research papers and categorized depending on the importance to user community. We conducted a survey on a software application defect prediction methodologies based on machine learning approaches as well as statistical approaches. This paper contains an outline of works that have been published so far and not a comprehensive review of all the papers published on the topic. We're confident that the survey of ours will help researchers to under- stand developments in this particular field of study in an effective and easy manner. We have also introduced as well as discussed the latest trends in defect prediction.