Landslide susceptibility prediction method and system based on semi-supervised support vector machine model

Abstract

The invention relates to a landslide susceptibility prediction method and system based on a semi-supervised support vector machine model. The method comprises the following steps of: acquiring an evaluation factor of a research area and a known landslide sample; calculating a deterministic coefficient value of each evaluation factor; carrying out susceptibility partitioning on the research area according to the deterministic coefficient value, and dividing the research area into five types of landslide susceptibility levels; selecting a non-landslide sample and known landslide sample to jointly form a first training test data set to perform training test on the support vector machine model; and adopting a trained and tested support vector machine model to predict an initial landslide susceptibility value of the research area, thereby determining an expanded landslide sample and a secondarily selected non-landslide sample to carry out training test on the support vector machine model, and further generating a semi-supervised support vector machine model to carry out landslide susceptibility prediction on the research area. According to the method, the landslide sample data acquisition process is simplified, and the precision of landslide susceptibility prediction is improved.

Description

technical field [0001] The invention relates to the technical field of geological disaster prediction, in particular to a landslide susceptibility prediction method and system based on a semi-supervised support vector machine model. Background technique [0002] The special topography, extreme climate and other factors lead to frequent landslide disasters, which often cause serious losses to the safety of local residents, building facilities and the environment. Using an appropriate evaluation model or method to calculate the possibility of landslide disasters, and then carry out landslide susceptibility evaluation, identify areas that are extremely prone to landslides, and reduce the adverse effects of landslides have important guiding significance for disaster prevention and mitigation. Therefore, it is necessary to strengthen the research on regional landslide susceptibility prediction to guide the disaster prevention and mitigation work in areas with high landslide occur...

AI Technical Summary

Problems solved by technology

On the one hand, although unsupervised machine learning does not require known landslide and non-landslide samples as model output variables in the training and testing process, due to the lack of prior knowledge guidance such as landslides and non-landslides, the modeling of unsupervised machine learning Accuracy is difficult to guarantee

On the other hand, landslide susceptibility prediction modeling based on fully supervised machine learning also has some deficiencies, mainly as follows: 1) It is difficult and expensive to conduct field surveys and obtain landslide sample data. It is generally difficult to obtain more complete samples of landslides; 2) The method of randomly selecting non-landslide samples in the entire research area during the modeling process brings a lot of errors to the training and testing process of fully supervised machine learning, which reduces the prediction of landslide susceptibility. the accuracy of

Images