Analysis of Slope Stability Based on Four Machine Learning Models
An Example of 188 Slopes
Abstract
To achieve rapid and precise prediction of slope stability, we propose an intelligent assessment method utilizing machine learning techniques. This approach aims to enhance the precision of slope stability evaluations, facilitating more effective and timely decision-making in geotechnical engineering. By analyzing 188 slope cases from domestic and international sources, we have identified six key feature variables to evaluate the Factor of Safety (FOS) for slope stability assessment. The dataset was established for evaluating slope stability, and to ensure robustness, it was divided into training and testing set using a 5-fold cross-validation approach. Four slope stability prediction models- GBM, SVM, XGB, and RF- were developed using machine learning algorithms. The accuracy of the models in predicting FOS for slopes was assessed using metrics such as MAE, MSE, RMSE, and R2. The best-performing machine learning model, along with the finite element model developed using GeoStudio, was applied to engineering examples to compare their feasibility and efficiency. The research findings demonstrates that the GBM model has a minimal error between the predicted and actual slope FOS, highlighting its high accuracy. The model shows a strong correlation between predicted and actual FOS, indicating its superior performance relative to other models. GBM model and the finite element model align well with the actual field conditions. However, the GBM model stands out due to its higher accuracy and faster computational efficiency. Therefore, the GBM model offers a high degree of fit between the predicted FOS and the actual values, making it well-suited for evaluating slope stability.
Keywords:
factor of safety, machine learning, slope stability, finite element modelPublished Online
How to Cite
Issue
Section
License
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (see The Effect of Open Access). As soon as the paper is accepted, finally submitted and edited, the paper will appear in the "OnlineFirst" page of the journal, thus from this point no other internet-based publication is necessary.

