What is it about?
Identifying landslide-prone areas can be approached as a machine learning classification problem. Various factors such as elevation, slope, aspect, terrain undulation, vegetation coverage, landform, and soil moisture content can be utilized as input variables for the classification algorithm to determine the probability of landslide occurrence within a specific range. Previous research in this area has been limited and fails to encompass diverse geological environments. This paper addresses this gap by utilizing a dataset collected within a province, which exhibits ample diversity and coverage. The dataset incorporates both continuous and discrete variables, allowing for a comprehensive evaluation of different classification algorithms. The primary objective of this study is to identify the most suitable classification algorithm among AdaBoost, cart, gbdt, xgboost, and random forest for those conditions.
Featured Image
Why is it important?
Geological disasters result in significant human and property losses. It is imperative to identify areas prone to geological disasters for prevention and monitoring purposes.
Perspectives
This study can be further used to establish the geological disaster prediction model. When the risk level can be more specific, the risk level can also be further accurately classified on the basis of this study.
Jingwen Zhou
Read the Original
This page is a summary of: Identification of High-Risk Areas for Geological Disasters using classification methods under complex environmental conditions, July 2023, ACM (Association for Computing Machinery),
DOI: 10.1145/3611450.3611458.
You can read the full text:
Contributors
The following have contributed to this page
