Seismic Vulnerability Assessment at an Urban Scale by Means of Machine Learning Techniques

Abstract

Seismic vulnerability assessment in urban areas would, in principle, require the detailed modeling of every single building and the implementation of complex numerical calculations. This procedure is clearly difficult to apply at an urban scale where many buildings must be considered; therefore, it is essential to have simplified, but at the same time reliable, approaches to vulnerability assessment. Among the proposed strategies, one of the most interesting concerns is the application of machine learning algorithms, which are able to classify buildings according to their vulnerability on the basis of training procedures applied to existing datasets. In this paper, machine learning algorithms were applied to a dataset which collects and catalogs the structural characteristics of a large number of buildings and reports the damage observed in L’Aquila territory during the intense seismic activity that occurred in 2009. A combination of a trained neural network and a random forest algorithm allows us to identify an opportune “a-posteriori” vulnerability score, deduced from the observed damage, which is compared to an “a-priori” vulnerability one, evaluated taking into account characteristic indexes for building’s typologies. By means of this comparison, an inverse approach to seismic vulnerability assessment, which can be extended to different urban centers, is proposed.