AN ESTIMATION MODEL FOR THE DEPTH OF ENGINEERING BEDROCK FROM TOPOGRAPHICAL DATA USING MACHINE LEARNING MODEL

Abstract

 A hazard map is a map showing the places where a natural disaster is likely to occur in the future. In the preparation of the map, the topography around the site is mentioned as a factor of large effect. For this topographical factor, the feature extraction by the rule base has been carried out by the expert arranging the causal relation between the disaster occurrence and the topography. However, considering the complexity of topographical information, these relationships cannot always be described as rules. From such situation, the approach by machine learning which automatically models the relation from the input and output data may become effective in the modeling which makes the topographical information to be an input. Then, in this study, the problem of estimating the depth of engineering bedrock from the topographical information was set as an example, and the effectiveness of the machine learning model was confirmed.