PREDICTION OF SLOPE FAILURE DURING THE 2018 HOKKAIDO EASTERN IBURI EARTHQUAKE BASED ON MACHINE LEARNING

Abstract

 The 2018 Hokkaido Eastern Iburi earthquake caused a large number of slope failures and casualties in Atsuma town and its surrounding areas. In previous studies, several prediction methods for slope failures have been proposed, and the prediction accuracy has been discussed. However, the prediction models suitable for extensive areas are not solidly developed. In view of the current situation, this study uses machine learning to predict the slope failures during the 2018 Hokkaido Eastern Iburi earthquake. In addition, the influence of NDVI of satellite image on the prediction results of machine learning model is discussed.