Classification of buildings' potential for seismic damage using a machine learning model with auto hyperparameter tuning

•The paper compares and evaluates the capability of various ML methods to classify the seismic damage potential of R/C buildings adequately.•The paper proposes a holistic system that automates the selection and application of the most appropriate algorithmic hyperparameters.•A large training dataset of R/C buildings analyzed for 65 real earthquake records using nonlinear dynamic analyses.•The SVM- Gaussian Kernel algorithm produced the highest classification results.•An auto hyperparameter tuning method for the winner algorithm is proposed, so that the hyperparameters are automatically optimized utilizing BO.