Der Einsatz Unterschiedlicher Klassifikationssysteme zur Rutschungserkennungs (in German)

A thunderstorm in Sachseln/Switzerland that had a rainfall total of 1997 and resulted in more than 700 soil slips. These soil slips were Federal Institute for Forest, Snow and Landscape Research and Germany. The results of the analysed soil slips highlighted that different factors, have a crucial influence on the slope stability. Besides the surrounding area were analysed and used to identify instable areas. data and creating new parameters that may have an influence on Linear Discriminant analysis (LDA), classification trees and Support describe the occurrence of soil slips were investigated, while at thereof. Analyses were conducted for the case where only cells of considered, as well as the case where the entire soil slip scars were strated that it is possible to recognize most of the soil slips (82 %), non-soil-slips) was low (75 %). After extensive research with different to non-soil-slip cells during the classification must be balanced. The results by 3 %. The use of the modern and time-consuming SVMs or classification trees did not result in better recognition rates.