Akademik Veri Yönetim Sistemi
9th International Conference on Electronics Computer and Computation (ICECCO 2012), Ankara, Türkiye, 1 - 03 Kasım 2012, ss.288-291, (Tam Metin Bildiri)
Software development projects require a critical and a costly testing phase to investigate the efficiency of the resultant product. As the size and complexity of the project increases, manual prediction of software defects becomes a time consuming and a costly task. An alternative to manual defect prediction is the use of automated predictors to focus on faulty modules and let the software engineer examine the defective part with more detail. In this aspect, improvement of fault predictors is an essential research topic in software quality projects. There are many base predictors tested or designed for this purpose. However, base predictors might be combined with the use of an ensemble strategy to improve their performance further, particularly their fault-detection abilities. The aim of this study is to demonstrate fault-prediction performance of ensemble predictors compared to baseline predictors empirically. In these experiments, we used 15 software projects from PROMISE repository to evaluate performance of the algorithms.