Testing machine learning models for heuristic building damage assessment applied to the Italian Database of Observed Damage (DaDO)

Ghimire, Subash; Guéguen, Philippe; Pothon, Adrien; Schorlemmer, Danijel

doi:https://doi.org/10.5194/nhess-23-3199-2023

Articles | Volume 23, issue 10

https://doi.org/10.5194/nhess-23-3199-2023

© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

Special issue:

Harmonized seismic hazard and risk assessment for Europe

https://doi.org/10.5194/nhess-23-3199-2023

© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 23, issue 10

Research article

|

05 Oct 2023

Research article |

| 05 Oct 2023

Testing machine learning models for heuristic building damage assessment applied to the Italian Database of Observed Damage (DaDO)

Subash Ghimire, Philippe Guéguen, Adrien Pothon, and Danijel Schorlemmer

Model code and software

Scikit-learn (https://scikit-learn.org/stable/install.html) F. Pedregosa, G. Varoquaux, L. Buitinck, G. Louppe, O. Grisel, and A. Mueller https://doi.org/10.1145/2786984.2786995

Short summary

This study explores the efficacy of several machine learning models for damage characterization, trained and tested on the Database of Observed Damage (DaDO) for Italian earthquakes. Reasonable damage prediction effectiveness (68 % accuracy) is observed, particularly when considering basic structural features and grouping the damage according to the traffic-light-based system used during the post-disaster period (green, yellow, and red), showing higher relevancy for rapid damage prediction.