Application of Fuzzy Linear Regression Modeling to Predict the Number of Program Faults

摘要

Software quality models can predict the quality of modules early enough for cost-effective prevention of problems. This paper introduces the Fuzzy Linear Regression Model, FLRM, modeling technique as a method for predicting fault ranges in software modules. FLRM differs from a classical linear regression model in that the output of FLRM is a fuzzy number. Predicting the exact number of faults in each program module is often not necessary. FLRM can predict the interval that the number of faults of each module falls into with a certain probability. A case study applies FLRM to a large military command, control, and communication system. We developed a FLRM using asymmetric triangular fuzzy coefficients to predict how likely the number of faults of each program will be in certain intervals based on software product metrics. We found that FLRM gives useful results.