The method neural networks of back propagation is discussed in this paper. Parameters of the original data for zoning and structure of the neural network are defined. It shows the results and assessments of accuracy landslide areas identification within Krasnaya Polyana. Proposal on the use of digital elevation models produced with high-precision geodetic techniques to improve the reliability of the simulation results is made.
This article is devoted to the problem of the information security of forecast in the study of landslides. The particular attention is paid to methods of mapping landslide risk areas and the organization of thematic database of the criteria of landslide danger. The using of GIS is offer for the solution of the problem of zoning on the degree of manifestation landslides as one of the most promising options.