Open AccessSelection of landslide affecting factors based on strong association analysis
Author(s) -
Luyao Li,
Rui Liu,
Xin Yang,
Mei Yang,
Yuantao Yang
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/780/7/072051
Subject(s) - landslide , support vector machine , selection (genetic algorithm) , computer science , data mining , apriori algorithm , machine learning , association rule learning , model selection , a priori and a posteriori , artificial intelligence , engineering , geotechnical engineering , philosophy , epistemology
The performance of models in landslide susceptibility mapping largely depends on the selection and combination of affecting factors. The Apriori algorithm proposed in this paper is a factor selection method of strong association analysis, which can select the factors that are most likely to induce landslides from 15 affecting factors. Then combine the origin and optimized factors to build the prediction model of landslide susceptibility by support vector machine (SVM) in machine learning. Afterwards, we verifying the landslide points in the dataset to measure the accuracy of the model. Ultimately, ROC curve was adopted to evaluate the prediction results of the two models. The result reveals that the model based on the combination of optimized factors (AUC 1=0.930) is superior to that based on 15 affecting factors (AUC 2=0.898).