基于SVM多类分类的滑坡区域危险性评价方法研究

STUDY ON LANDSLIDE HAZARD ZONATION BASED ON MULTI-CLASSIFICATION SUPPORT VECTOR MACHINE

近年来,随着新理论、新技术得发展,提出了许多新模型和方法应用于滑坡区域危险性评价中。支持向量机(support vector m ach ine,SVM)是新一代的学习算法,已有前人利用SVM应用于滑坡灾害预测中。然而大多只是利用了SVM的两分类算法,得到的结果只有稳定不稳定两种,这对滑坡区域评价来说是远远不够的。本文尝试利用SVM的多类分类算法进行滑坡危险性区域评价,取得了较好的结果。

Many new models have been applied to landslide hazard zonation since 1990s. Support vector machine （SVM） as a novel learning algorithm has been used in the fields of geohazards, but in most of them, c-SVM algorithm is used to make sample data into two classification. This paper uses multi-classification algorithm of support vector machine to produce landslide hazard zonation figure and the evaluation results indicate that it is effective and reasonable.