Software cost estimation using machine learning algorithms

Abstract

Software cost estimation is one of the most important problems in software projects. This chapter focuses on the estimation cost of software projects by testing different machine learning algorithms using the Waikato Environment for Knowledge Analysis (WEKA) data mining software tool. The estimation process includes size estimation, effort estimation, development of initial project schedules and, finally, estimation of the overall project cost. Algorithms were applied to a Chinese dataset taken from the PROMISE data repository. The WEKA contains a large number of machine learning algorithms for data preprocessing, clustering, classification, regression, visualization and feature selection. A total of 29 classification algorithms in the WEKA were applied to the Chinese dataset. The chapter presents the performance evaluation results of the machine learning algorithms applied to the Chinese dataset. The analysis of the test results showed that the best prediction algorithm in the Chinese dataset was the SMOreg algorithm. © ISTE Ltd 2022.