An Application of SVM-Based Classification in Landslide Stability. Issue 2 (2nd April 2016)

• Record Type: Journal Article
• Authors: Jiang, Tingyao; Lei, Peng; Qin, Qin
• Is Part Of: Intelligent automation & soft computing. Volume 22:Issue 2(2016)
• Page Start: 267
• Page End: 271
• Publication Date: 2016-04-02
• Subjects: Landslide -- Stability evaluation -- SVM; Artificial intelligence -- Periodicals; Intelligent control systems -- Periodicals; 003.5
• DOI: 10.1080/10798587.2015.1095480
• Languages: English
• ISSNs: 1079-8587

Abstract

Abstract: The calculation method of landslide stability is a critical issue in landslide research. SVM-based multi-classification algorithm, which can structure multiple binary classifiers to accomplish the multi-classification task is used for landslide stability analysis. In this paper, the slope height, slope angle, capacity, internal friction angle and cohesion are selected as impact factors affecting the stability of landslide. Loop crossover method is used to verify the accuracy of the algorithm. Compared with the Mahalanobis distance and Bayes discriminant, the proposed algorithm has a better prediction result, but it also has the largest mis-judgment loss. The accuracy of Bayes discriminant is less than the SVM, but its mis-judgment loss is minimal.