Defect Prediction Framework using Neural Networks for Business Intelligence Technology Based Projects

So far, researchers in field of defect prediction have published multiple approaches, but none of these publications have identified the Business Intelligence project life cycle. In this paper, we have taken data from 60 BI projects from a Data Analytics organization and have used the real project data to design a prediction model based on artificial neural networks. Results are evaluated with comparison of three different training algorithms, i.e., Lavenberg-Marquardt, Scaled Conjugate Gradient and Bayesian Regularization backpropagation algorithms, in the view of their ability to perform defect prediction. The objective is to design a prediction framework, which is expected to be effective and acceptable for predicting the defects in multiple phases across Business Intelligence projects.