Assessment of Landslide Causal Factors Using ANN Method

Abstract

In this study landslide causal factors which are considered to have the same effect in assessment techniques are categorized and their impact on landslides is analyzed to acquire reasonable weighting factors in the landslide hazard.Results are compared to those of the Assessment Chart developed by National Institute for Disaster Prevention (NIDP) and the adequacy and proper portion for landslide causal factors are considered. The Artificial Neural Network (ANN) method applied to 28 landslide areas is incorporated to evaluate the reasonable rating. Results show that the following items in the Chart are necessary to modify their portions in order to implement the precise assessment results: 1) Estimated damage; 2) Tension crack; 3) Existence of valley.