An Efficient Software Fault Prediction Model using Cluster based Classification

Predicting fault -prone software components is an economically important activity due to limited budget allocation for software testing. In recent years data mining techniques are used to predict the software faults .In this research, we present a cluster based fault prediction classifiers which increases the probability of detection. The expectation from a predictor is to get very high probability of detection to get more reliable and test effective software. In our experiments, we used fault data from mission critical systems. In this paper we have used discretization as preprocessing and cluster based classification for prediction of fault-prone software modules. Clustering based classification allows production of comprehensible models of software faults exploiting symbolic learning algorithms. To evaluate this approach we perform an extensive comparative analysis with benchmark results of software fault prediction for the same data sets. Our proposed model shows better results than the standard and benchmark approaches for software fault prediction. Our proposed model gives superior probability of detection (pd) 83.3% and balance rates 685%.