AN EMPIRICAL STUDY ON REGRESSION ANALYSIS TO PREDICT SOFTWARE EFFORT BASED ON USE CASE POINTS
International Journal of Latest Trends in Engineering and Technology · 2018
The effort estimation of software development is a process of predicting the most realistic effort (person-hours) required to maintain and development of software.This paper aims to provide best regression technique among linear regression, multi linear regression, exponential regression, and power regression to predict software effort using use case points.The main contribution of this study is a comparative analysis of different regression techniques.To validate the accuracy of regression techniques we have used R 2 value, MMRE and TSS.From this study we observed that linear regression, exponential regression, and power regression produced same results with some variance over multi linear regression.