Conceptual cost estimates, evaluated during a construction project's initiation phase, are fundamental for determining whether to invest in the project, validating its budget, or screening alternatives. Compared to traditional estimation techniques, artificial intelligence (AI) methods proved effective in assessing the nonlinear relationship between project variables and actual cost at completion. Due to the number and variability of available studies, it is not clear which AI techniques are most effective. This study systematically reviews previous works employing AI for conceptual cost estimation, focusing on the techniques adopted and the scorers utilised. The results show a rising trend in AI adoption, including supervised machine learning, knowledge-based, and evolutionary techniques. Performance-wise, the results hint at gradient boosting, random forest, and neural networks proving superior to both genetic algorithms and case-based reasoning techniques, which in turn prove superior to linear models. This review provides a brief overview of possible AI techniques and performance scorers to utilise for conceptual cost estimation in construction projects.
A BRIEF REVIEW OF ARTIFICIAL INTELLIGENCE TECHNIQUES FOR CONCEPTUAL COST ESTIMATION IN CONSTRUCTION PROJECTS / Ottaviani, F. M.; De Marco, A.; Audisio, B.; Wong, J.; Belack, C.. - (2024), pp. 615-624. (Intervento presentato al convegno 40th Annual Conference on Association of Researchers in Construction Management, ARCOM 2024 tenutosi a London (UK) nel 2024).
A BRIEF REVIEW OF ARTIFICIAL INTELLIGENCE TECHNIQUES FOR CONCEPTUAL COST ESTIMATION IN CONSTRUCTION PROJECTS
Ottaviani F. M.;De Marco A.;
2024
Abstract
Conceptual cost estimates, evaluated during a construction project's initiation phase, are fundamental for determining whether to invest in the project, validating its budget, or screening alternatives. Compared to traditional estimation techniques, artificial intelligence (AI) methods proved effective in assessing the nonlinear relationship between project variables and actual cost at completion. Due to the number and variability of available studies, it is not clear which AI techniques are most effective. This study systematically reviews previous works employing AI for conceptual cost estimation, focusing on the techniques adopted and the scorers utilised. The results show a rising trend in AI adoption, including supervised machine learning, knowledge-based, and evolutionary techniques. Performance-wise, the results hint at gradient boosting, random forest, and neural networks proving superior to both genetic algorithms and case-based reasoning techniques, which in turn prove superior to linear models. This review provides a brief overview of possible AI techniques and performance scorers to utilise for conceptual cost estimation in construction projects.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/11583/2994306
Attenzione
Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo