Improving the prediction of ground motion parameters based on an efficient bagging ensemble model of M5′ and CART algorithms

In the present study, an efficient bagging ensemble model based on two well-known decision tree algorithms, namely, M5′ and Classification and Regression Trees (CART) is utilized so as to estimate the peak time-domain strong ground motion parameters. Four different predictive models, namely, CART, Ensemble M5′, Ensemble CART, and Ensemble M5′ + CART are developed to evaluate Peak Ground Acceleration, Peak Ground Velocity, and Peak Ground Displacement. A big database from the Pacific Earthquake Engineering Research Center is employed so as to develop the proposed models. Earthquake magnitude, earthquake source to site distance, average shear-wave velocity, and faulting mechanisms are considered as the predictive parameters. The superior performances of the developed models are observed in the validation against the most recent soft computing based models available in the specialized literature. Parametric as well as sensitivity analyses are carried out to ensure the robustness of the predictive models in discovering the physical concept latent in the nature of the problem.