A Software Project Estimation Method Based on Decision Trees
Wei Yan · Computer Engineering and Science · 2009
An important problem in the field of software engineering is to make predictions concerning the size effort/cost of software development projects,namely software project estimation. The prediction methods based-on machine learning are more effective than other ones,and this paper presents an estimation method based-on decision trees,which is one of the commonly used machine learning tools. This technique predicts the value scope of the target project by classifying the 'effort' attributes. We exam the use of decision trees in practice,and the results from a subset of the ISBSG v9 data set show that the demands of software project estimation are met.