ABSTRACT
This chapter aims at evaluating the results of GIS based landslide susceptibility mapping by two methods: the analytical hierarchy process and the logistic regression model. It argues that the logistic regression map was better in matching the historical landslide density than the analytical hierarchy process method. Applications of bivariate statistical landslide susceptibility assessment methods have been widly reported. In the application of logical regression model for landslide susceptibility mapping, some researchers have tried to exploit the data via using dummy binary variables. Extensive field studies were used to check the size and the shape of landslides; and to identify the type of movements and the materials involved. All training data were input to the logistic regression algorithm within statistical product and service solutions and IDRISI Kilimanjaro to calculate the correlation of landslide.
