An empirical evaluation of ensemble adjustment methods for analogy-based effort estimation

TL;DR

This study empirically evaluates ensemble learning techniques for analogy-based effort estimation, demonstrating that ensembles can improve accuracy over single methods, especially when using linear adjustment variants.

Contribution

It provides a large-scale comparison of ensemble adjustment methods, highlighting their effectiveness and recommending linear adjustment ensembles for better performance.

Findings

Ensemble methods significantly improve predictive accuracy.

Linear adjustment ensembles outperform other variants.

Ensembles are often better than single adjustment methods.

Abstract

Objective: This paper investigates the potential of ensemble learning for variants of adjustment methods used in analogy-based effort estimation. The number k of analogies to be used is also investigated. Method We perform a large scale comparison study where many ensembles constructed from n out of 40 possible valid variants of adjustment methods are applied to eight datasets. The performance of each method was evaluated based on standardized accuracy and effect size. Results: The results have been subjected to statistical significance testing, and show reasonable significant improvements on the predictive performance where ensemble methods are applied. Conclusion: Our conclusions suggest that ensembles of adjustment methods can work well and achieve good performance, even though they are not always superior to single methods. We also recommend constructing ensembles from only linear adjustment methods, as they have shown better performance and were frequently ranked higher.