An agile effort estimation based on story points using machine learning techniques

Abstract

Nowadays, many software companies face the problem of predicting the accurate software effort. Most of the software projects are failed due to over budget and over schedule as well as under-budget and under-schedule. The main reason for the failure of software projects is inaccurate effort estimation. To improve the accuracy of effort estimation, various effort estimation techniques are introduced. Functional points, object points, use case points, story points, etc., are used for effort estimation. Earlier, traditional process models like waterfall model, incremental model, spiral model, etc., are used for developing the software, but none of them have given the successful projects to the customers. Now, 70% of the application software’s have been developed by the agile approaches. The success rate of the projects developed by using agile methodologies has been increased. The major objective of this research is to estimate the effort in agile software development using story points. The obtained results have been optimized using various Machine Learning Techniques to achieve an accurate prediction of effort and compared performance measures like MMRE, MMER, and PRED.