24th EANN 2023, 14 - 17 June 2023, León, Spain

A Machine Learning Approach for Seismic Vulnerability Ranking

Ioannis Karampinis, Lazaros Iliadis


  Structures often suffer damages as a result of earthquakes, potentially threatening human lives, disrupting the economy and requiring large amounts of monetary reparations. Thus, it is essential for governments to be able to rank a given population of structures according to their expected degree of damage in an earthquake, in order for them to properly allocate the available resources for prevention. In this paper, the authors present a ranking approach, based on Machine Learning (ML) algorithms for pairwise comparisons, coupled with ad hoc ranking rules. The degree of damage of several structures from the Athens 1999 earthquake, along with collected attributes of the building, were used as input. The performance of the ML classification algorithms was evaluated using the respective metrics of Precision, Recall, F1-score, Accuracy and Area Under Curve (AUC). The overall performance was evaluated using Kendall's tau distance and by viewing the problem as a classification into bins. The obtained results were promising, outperforming currently employed engineering practices. They have shown the capabilities and potential of these models in mitigating the effects of earthquakes on society.  

*** Title, author list and abstract as seen in the Camera-Ready version of the paper that was provided to Conference Committee. Small changes that may have occurred during processing by Springer may not appear in this window.