Computational Technology Resources - CCC

E.K. Nyarko¹, S. Czarnecki², A.D.R. Troncoso Garcia³, M. Hadzima-Nyarko⁴, F. Martinez Alvarez³, B. Bulajić⁵, E. Isik⁶ and D. Radu⁷

¹Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, Josip Juraj Strossmayer University of Osijek, Croatia
²Department of Materials Engineering and Construction Processes, Wroclaw University of Science and Technology, Wrocław, Polska
³Data Science & Big Data Lab, Pablo de Olavide University, Spain
⁴Faculty of Civil Engineering and Architecture, Josip Juraj Strossmayer, University of Osijek, Croatia
⁵Faculty of Technical Sciences, University of Novi Sad, Serbia
⁶Department of Civil Engineering, Bitlis Eren University, Turkey
⁷Faculty of Civil Engineering, Transilvania University of Braşov, Romania

doi:10.4203/ccc.14.3.1

E.K. Nyarko, S. Czarnecki, A.D.R. Troncoso Garcia, M. Hadzima-Nyarko, F. Martinez Alvarez, B. Bulajić, E. Isik, D. Radu, "Predicting Earthquake-Induced Building Damage Grades Using Machine Learning and Explainable AI", in P. Iványi, J. Kruis, B.H.V. Topping, (Editors), "Proceedings of the Sixteenth International Conference on Computational Structures Technology", Civil-Comp Press, Edinburgh, UK, Online volume: CCC 14, Paper 3.1, 2026, doi:10.4203/ccc.14.3.1

Keywords: building damage grade, machine learning, seismic performance, PGA, random forest, neural network, gradient boosting.

Abstract

Machine learning techniques are increasingly being applied in civil engineering to improve the assessment of structural performance and post-earthquake damage evaluation. In this study, several machine learning approaches, including regression models, classification models, and Artificial Neural Networks, were investigated for the prediction of seismic building damage grades using a database composed of structural, geometrical, and seismic parameters collected after major earthquakes in Croatia and Türkiye. Different data stratification strategies were considered to evaluate the robustness and predictive capability of the developed models. The results demonstrated that machine learning methods can effectively capture complex nonlinear relationships associated with seismic damage assessment and provide accurate prediction of building damage grades. Among the investigated approaches, ensemble-based classification models and ANN architectures achieved the most favorable predictive performance depending on the adopted stratification strategy. To improve model transparency and engineering interpretability, SHapley Additive exPlanations (SHAP) analysis was additionally performed to investigate the influence of individual input variables on the predicted damage grades. The obtained results confirmed the potential of machine learning and explainable artificial intelligence techniques as efficient tools for rapid seismic damage assessment and decision support in post-earthquake engineering applications.

download the full-text of this paper (PDF, 13 pages, 309 Kb)

go to the previous paper
go to the next paper
return to the table of contents
return to the volume description

	Computational & Technology Resources an online resource for computational, engineering & technology publications
	not logged in - login
Front Page Browse CCC CCP CSETS CTR IJRT Other Authors Search Purchase Guide FAQ Contact us	Civil-Comp Conferences ISSN 2753-3239 CCC: 14 PROCEEDINGS OF THE SIXTEENTH INTERNATIONAL CONFERENCE ON COMPUTATIONAL STRUCTURES TECHNOLOGY Edited by: P. Iványi, J. Kruis and B.H.V. Topping Paper 3.1 Predicting Earthquake-Induced Building Damage Grades Using Machine Learning and Explainable AI E.K. Nyarko¹, S. Czarnecki², A.D.R. Troncoso Garcia³, M. Hadzima-Nyarko⁴, F. Martinez Alvarez³, B. Bulajić⁵, E. Isik⁶ and D. Radu⁷ ¹Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, Josip Juraj Strossmayer University of Osijek, Croatia ²Department of Materials Engineering and Construction Processes, Wroclaw University of Science and Technology, Wrocław, Polska ³Data Science & Big Data Lab, Pablo de Olavide University, Spain ⁴Faculty of Civil Engineering and Architecture, Josip Juraj Strossmayer, University of Osijek, Croatia ⁵Faculty of Technical Sciences, University of Novi Sad, Serbia ⁶Department of Civil Engineering, Bitlis Eren University, Turkey ⁷Faculty of Civil Engineering, Transilvania University of Braşov, Romania doi:10.4203/ccc.14.3.1 Full Bibliographic Reference for this paper E.K. Nyarko, S. Czarnecki, A.D.R. Troncoso Garcia, M. Hadzima-Nyarko, F. Martinez Alvarez, B. Bulajić, E. Isik, D. Radu, "Predicting Earthquake-Induced Building Damage Grades Using Machine Learning and Explainable AI", in P. Iványi, J. Kruis, B.H.V. Topping, (Editors), "Proceedings of the Sixteenth International Conference on Computational Structures Technology", Civil-Comp Press, Edinburgh, UK, Online volume: CCC 14, Paper 3.1, 2026, doi:10.4203/ccc.14.3.1 Keywords: building damage grade, machine learning, seismic performance, PGA, random forest, neural network, gradient boosting. Abstract Machine learning techniques are increasingly being applied in civil engineering to improve the assessment of structural performance and post-earthquake damage evaluation. In this study, several machine learning approaches, including regression models, classification models, and Artificial Neural Networks, were investigated for the prediction of seismic building damage grades using a database composed of structural, geometrical, and seismic parameters collected after major earthquakes in Croatia and Türkiye. Different data stratification strategies were considered to evaluate the robustness and predictive capability of the developed models. The results demonstrated that machine learning methods can effectively capture complex nonlinear relationships associated with seismic damage assessment and provide accurate prediction of building damage grades. Among the investigated approaches, ensemble-based classification models and ANN architectures achieved the most favorable predictive performance depending on the adopted stratification strategy. To improve model transparency and engineering interpretability, SHapley Additive exPlanations (SHAP) analysis was additionally performed to investigate the influence of individual input variables on the predicted damage grades. The obtained results confirmed the potential of machine learning and explainable artificial intelligence techniques as efficient tools for rapid seismic damage assessment and decision support in post-earthquake engineering applications. download the full-text of this paper (PDF, 13 pages, 309 Kb) go to the previous paper go to the next paper return to the table of contents return to the volume description
Back to top	©Civil-Comp Limited 2023 - terms & conditions