Statistical modeling of tree failures during storms

摘要: Abstract The failure of trees during storms imposes strong economic and societal costs. Statistical modeling for predicting the probability a tree failing has potential to help improve risk management. purpose this study is explore predictability using advanced predictive approach. These models also have broader applicability failures technical systems adverse weather events. To train test models, we use data set from real case in Massachusetts, USA. We compare out-of-sample accuracy several machine learning including logistic regression, classification regression trees, multivariate adaptive splines, artificial neural network, naive-Bayes random forest, boosting, an ensemble model boosting forest. Our results demonstrate that forest achieves best prediction storm. can care professionals make better decisions reduce prior