Landslide Prediction in Three Gorges Based on Cloud Model and Data Field

Landslide spatial prediction in Three Gorges is studied based on spatial data mining. China-Brazil earth resources satellite (Cbers) image, 1: 0.05 million geological graph and 1:0.01 million relief map were adopted as the data origin to produce the key factors influencing landslide transmutation. A spatial data mining algorithm is proposed which is suitable for landslide spatial prediction. The algorithm adopts data field to synthetically analyze the spatial distribution of landslides and the key factors influencing landslide transmutation and extract the potential centers. Then the concept represented by each potential center is described by a cloud model and elevated by the synthesized cloud method to produce the high-level concept. Clustering analysis is made according to the membership degree of each data point to each high-level concept, and to realize the landslide spatial prediction in Three Gorges. The experimental results have shown that the algorithm proposed in the paper possesses a high prediction precision and the prediction result is obviously priori to the ones of the 4 methods of IsoData, k-means, parallelepiped and minimum distance. So the algorithm can effectively realize landslide spatial prediction and is a suitable data mining algorithm for landslide prediction.