Journal cover Journal topic
Proceedings of the ICA
Journal topic
Articles | Volume 4
Proc. Int. Cartogr. Assoc., 4, 67, 2021
https://doi.org/10.5194/ica-proc-4-67-2021
Proc. Int. Cartogr. Assoc., 4, 67, 2021
https://doi.org/10.5194/ica-proc-4-67-2021

  03 Dec 2021

03 Dec 2021

Landslide susceptibility evaluation based on optimized support vector machine

Jiping Liu1, Rongfu Lin1,2, Shenghua Xu1, Yong Wang1, Xianghong Che1, and Jie Chen1 Jiping Liu et al.
  • 1Chinese Academy of Surveying and Mapping, China
  • 2School of Geomatics, Liaoning Technology University, China

Keywords: evaluation of landslide susceptibility, support vector machine

Abstract. Landslide is a natural disaster that has caused great property losses and human casualties in the world. To strengthen the target prevention and management level, ZhaShui county, Shaanxi province, is selected as the research area to evaluate the landslide susceptibility. First of all, under the premise of considering the correlation, 10 evaluation factors closely related to landslide disaster (i.e., elevation, rainfall, rock group, slope, slope aspect, vegetation index, landform, distance to residential area, distance to road, distance to river system) are taken together with non-landslide points, which are selected under multi-constraint conditions to form a sample data-set. Secondly, the sample dataset is substituted into the Support Vector Machine (SVM) model optimized by firefly algorithm for training and prediction. Finally, the result map was partitioned according to the natural discontinuous point method, and the landslide susceptibility map was obtained. The results show that the model optimized by the firefly algorithm has higher accuracy, and the landslide susceptibility results are more consistent with the actual distribution of disaster points.

Publications Copernicus
Download
Citation