Software Fault Prediction by Linear Regression Using Factor Analysis

Abstract

Software practitioners develop models by considering the process of software fault prediction in the early stage of the software development life cycle (SDLC) in order to detect faulty classes or modules. Various statistical and machine learning techniques were examined in the past for software fault prediction. newlineAn empirical analysis of Chidamber and Kemerer (CK) and object oriented (OO) metrics with review of studies from year 1996 to 2018 in the literature considering the statistical and machine learning techniques for software fault prediction is presented in the work. On a set of benchmark data, for software faults and metrics to identify the underlying latent variables models of fault prediction based on multiple linear regression on the newlineidentified factors are proposed. newlineThe results obtained establish the potential and capabilities of the factor analysis for grouping important factors and using regression to identify the significant predictors. However, the significance and the application of the factor analysis with regression in software fault prediction are still limited and further studies should be considered in order to generalize the results.