Data-Driven Construction Cost Forecasting: Integrating Big Data Analytics and Machine Learning Models

Accurate construction cost forecasting is essential for effective project decision-making and cost control throughout the project lifecycle. Traditional cost estimation methods, which are largely based on historical data analysis, parametric models, and expert judgment, often struggle to cope with the complexity and uncertainty of contemporary construction projects. With the increasing adoption of digital technologies such as Building Information Modeling (BIM), project management information systems, and digital procurement platforms, large volumes of cost-related data are continuously generated, creating new opportunities for data-driven cost forecasting.This paper explores the application of big data analytics and machine learning techniques in construction cost forecasting and proposes an integrated data-driven framework that links data preparation, model development, and managerial decision-making. Rather than focusing on a single forecasting algorithm, the study emphasizes the importance of aligning data characteristics, modeling approaches, and practical cost management contexts. Representative machine learning models, including neural networks, support vector machines, and tree-based methods, are discussed and compared in terms of their predictive capability, data requirements, and interpretability.