Artificial Intelligence and Parametric Construction Cost Estimate Modeling: State-of-the-Art Review

This study reviews the common practices and procedures conducted to identify the cost drivers that the past literature has classified into two main categories: qualitative and quantitative procedures. In addition, the study reviews different computational intelligence (CI) techniques and ensemble methods conducted to develop practical cost prediction models. This study discusses the hybridization of these modeling techniques and the future trends for cost model development, limitations, and recommendations. The study focuses on reviewing the most common artificial intelligence (AI) techniques for cost modeling such as fuzzy logic (FL) models, artificial neural networks (ANNs), regression models, case-based reasoning (CBR), hybrid models, diction tree (DT), random forest (RF), supportive vector machine (SVM), AdaBoost, scalable boosting trees (XGBoost), and evolutionary computing (EC) such as genetic algorithm (GA). Moreover, this paper provides the comprehensive knowledge needed to develop a reliable parametric cost model at the conceptual stage of the project. Additionally, field canals improvement projects (FCIPs) are used as an actual case study to analyze the performance of the ML models. Out of 20 AI techniques, the results showed that the most accurate and suitable method is XGBoost with 9.091% and 0.929 based on mean absolute percentage error (MAPE) and adjusted R2, respectively. Nonlinear adaptability, handling missing values and outliers, model interpretation, and uncertainty are discussed for the 20 developed AI models. In addition, this study presents a publicly open data set for FCIPs to be used for future model validation and analysis.