Deep Learning Model for Function Point Based Software Cost Estimation -An Industry Case Study

Cost estimation of software engineering is a vital mission for software development organization. Accurate estimation can be valuable for both the developers and the customers. How to make estimation more accurate and reproducible is an important task to be settled in both research and industry area. This paper gives a review of the state of the art in the field of software cost estimation and proposes a Deep Neural Network(DNN) model to facilitate the usage of function point estimation. The results show that the function point based estimation can be a good choice in the early evaluation of the software budget while the deep neural network can significantly improve the labor productivity and the accuracy of the evaluation result .