Predicting High-Rise Buildings Construction Cost Using Machine Learning-Based Ensemble Model

This study investigates the application of machine learning (ML)-based ensemble models for predicting the construction cost of high-rise buildings. Three ensemble models - Voting, Bagging, and Stacking - were developed from three individual ML models, including artificial neural networks (ANNs), support vector regression (SVR), and classification and regression trees (CART). The analytical results demonstrate that ensemble models achieve higher predictive accuracy than individual models. The ANNs+SVR voting ensemble emerged as the most effective predictive model, followed by the ANNs+SVR+CART voting ensemble. These findings highlight the potential of ML-based ensemble techniques as a practical and reliable tool for project owners and contractors to enhance cost estimation accuracy in high-rise building construction projects.