AI approaches in earthquake damage prediction: A study on RC buildings with logistic regression

This study aims to determine the earthquake performance of reinforced concrete structures in Dogubayazıt district of Agri province, which is located in a region with high seismic activity according to Turkey's Earthquake Hazard Map, and to predict the potential damage these structures may incur during an earthquake.

The report states that the key risk factors contributing to building damage include pounding effects, which significantly increase the likelihood of moderate or severe damage, and soft-story configurations, with many of these structures predicted to sustain severe damage. Heavy overhangs and irregular building plans further exacerbate the risks, with approximately 80% of such structures expected to experience moderate or severe damage. Conversely, buildings with robust structural elements such as shear walls, columns, and beams are projected to be more resilient, sustaining only minor damage.