A Comparison Between Decision Trees and Decision Tree Forest Models for Software Development Effort Estimation

摘要

Accurate software effort estimation has been a challenge for many softwarepractitioners and project managers. Underestimation leads to disruption in theprojects estimated cost and delivery. On the other hand, overestimation causesoutbidding and financial losses in business. Many software estimation modelsexist; however, none have been proven to be the best in all situations. In thispaper, a decision tree forest (DTF) model is compared to a traditional decisiontree (DT) model, as well as a multiple linear regression model (MLR). Theevaluation was conducted using ISBSG and Desharnais industrial datasets.Results show that the DTF model is competitive and can be used as analternative in software effort prediction.