A Study of Data Pre-Processing Techniques for Machine Learning Algorithm to Predict Software Effort Estimation

Abstract

Software cost estimation requires high precision, however exact estimations are hard to accomplish. Progressively, data mining is utilized to enhance an association's product process quality, e.g. the exactness of cost estimations. There are countless strategies exists for software cost estimation, selecting the most suitable and pre-processing of data that is used for machine learning based software cost estimation is the subject of this paper. The effort invested in a software project is probably one of the most important and most analysed variables in recent years in the process of project management. Software effort estimates is an important part of software development work and provides essential input to project feasibility analyses, bidding, budgeting and planning. Analogy-based estimates models emerge as a promising approach, with comparable accuracy to arithmetic methods, and it is potentially easier to understand and apply. Studies show all the models are sensitive to the quality and availability data, thus requiring a systematic data treatment. In this paper we investigate various methods of data pre-processing for machine learning techniques based software cost estimation.