Scholarlink

Machine learning models are implemented to predict the factor of safety (FOS) and slope stability using Python in Google Colab. For this purpose, 113 global dry slope cases are collected. The key parameters are optimized through fivefold cross-validation, and the entire data is divided into test and train with an 80:20 split. Different models like Gaussian Naive Bayes (NB), random forest (RF), K-nearest neighbor (KNN), and decision tree classifier (DC) are applied successfully as classification models, whereas KNN, linear regression (LR), RF, and xtreme gradient boost (XGB) as regression models. The performance of these models is evaluated based on precision, f1-score, recall value, and accuracy in the case of the classification model, whereas mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and R-squared values are used for the regression models. The results are compared, and RF is revealed as the best-suited model in both cases having the highest accuracy value and lowest error rate.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Slope Stability Prediction Using Machine Learning Algorithms as Classification Models and Regression Models

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Sandipta Choudhury,
  • Debabrata Giri

摘要

Machine learning models are implemented to predict the factor of safety (FOS) and slope stability using Python in Google Colab. For this purpose, 113 global dry slope cases are collected. The key parameters are optimized through fivefold cross-validation, and the entire data is divided into test and train with an 80:20 split. Different models like Gaussian Naive Bayes (NB), random forest (RF), K-nearest neighbor (KNN), and decision tree classifier (DC) are applied successfully as classification models, whereas KNN, linear regression (LR), RF, and xtreme gradient boost (XGB) as regression models. The performance of these models is evaluated based on precision, f1-score, recall value, and accuracy in the case of the classification model, whereas mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and R-squared values are used for the regression models. The results are compared, and RF is revealed as the best-suited model in both cases having the highest accuracy value and lowest error rate.