An Evaluation of Software Cost Estimation Techniques: From Expert Analysis to Algorithmic Approaches

Abstract

The estimation of software costs is a critical element of project management, as it aids in the planning of budgets, the allocation of resources, and the scheduling of projects. This research work presents a comparative analysis of software cost estimation methodologies. The primary areas of investigation in the paper are expert judgment and algorithmic models. We provide a thorough examination of the numerous expert judgment methods, including the most common approaches and the associated advantages and disadvantages. Additionally, algorithmic models undergo analyses, which encompass sophisticated machine learning algorithms and regression-based methodologies. The associated algorithms' advantages and disadvantages are underscored in these analyses. A comparative study is conducted using predetermined criteria to provide insight into the relative efficacy of algorithmic models in comparison to expert judgment.