M.: Improved Bayesian Networks for Software Project Risk Assessment Using Dynamic Discretisation, Software Engineering Techniques: Design for Quality

Abstract. It is possible to build useful models for software project risk assess-ment based on Bayesian networks. A number of such models have been pub-lished and used and they provide valuable predictions for decision-makers. However, the accuracy of the published models is limited due to the fact that they are based on crudely discretised numeric nodes. In traditional Bayesian network tools such discretisation was inevitable; modelers had to decide in ad-vance how to split a numeric range into appropriate intervals taking account of the trade-off between model efficiency and accuracy. However, recent a recent breakthrough algorithm now makes dynamic discretisation practical. We apply this algorithm to existing software project risk models. We compare the accu-racy of predictions and calculation time for models with and without dynamic discretisation nodes.