Random Forests for Predicting Software Effort Estimation Based on Use-Case Points Analysis

EasyChair Preprint 9439

Abstract

Software estimation is vital for the success of software engineering projects. However, predicting software effort is difficult due to the complexity, intangibility, and diversity of software solutions and its involved expertise and underlying technology. This paper aims at enhancing the accuracy of software estimation using a data mining approach that combines Random Forests Regression with Use-Case Points analysis that is typically used in estimating effort in object-oriented software engineering projects. The experimental results of applying our proposed approach have demonstrated a significant improvement in the prediction accuracy of software effort estimation when compared to Use-Case Points estimation based on R2 and other metrics such as MAE, and MSE.

Keyphrases: Data Mining, Random Forests, Software effort estimation, Use Case Points, machine learning

@booklet{EasyChair:9439,
  author    = {Ne'Mah Alrababa'H and Ahmed Banimustafa},
  title     = {Random Forests for Predicting Software Effort Estimation Based on Use-Case Points Analysis},
  howpublished = {EasyChair Preprint 9439},
  year      = {EasyChair, 2022}}