Abstract

Construction cost estimation remains a critical challenge in project management, with traditional methods often lacking accuracy and efficiency. This paper presents a comprehensive analysis of machine learning (ML) approaches for construction cost estimation, comparing Random Forest, Support Vector Regression, Gradient Boosting, and Neural Network models. Through a case study of 2,847 residential construction projects, The research demonstrates that ensemble methods achieve superior performance with Random Forest attaining 92.3% accuracy and 8.7% MAPE. The findings indicate ML models significantly outperform traditional parametric estimation methods, offering improved accuracy and reduced estimation time from weeks to minutes.