|
| Integration of weighted Rock Mass quality Evaluation method based on improved Machine Learning algorithm based on RandomOverSampler |
|
Received:October 19, 2023
Revised:November 07, 2023
Accepted:December 12, 2023
Published Online:March 24, 2025
|
| View Full Text View/Add Comment Download reader |
| DOI： |
| KeyWord:ROS algorithm;machine learning algorithm; improved entropy weight integration method |
| Author | Institution |
| CHENG yuyin |
Yunnan Phosphate Haikou Co Ltd;China;Faculty of Land Resources Engineering,Kunming University of Science and Technology;China |
| LIU Jian |
Faculty of Land Resources Engineering,Kunming University of Science and Technology |
|
| Hits: 411 |
| Download times: 174 |
| Abstract: |
| Rock mass quality evaluation is a qualitative method to evaluate rock mass quality, and accurate rock mass classification is a difficult problem in deep underground engineering. In order to solve the problem of misjudgment in a single rock mass evaluation model, an integrated weighted rock mass quality evaluation method based on ROS improved machine learning algorithm is proposed. The unbalanced samples are oversampled by ROS algorithm, and the preliminary classification results are obtained by SVM, KNN, naive Bayesian, decision tree, MLP, RBF, RandomForest, linear discrimination, LightGBM, XGBoost and GradientBoosting. Taking each classifier as an index, the weight of each classifier is determined, the integrated weight rock mass quality evaluation model of ROS and machine learning algorithm is established, and the comprehensive discrimination result is obtained, which greatly reduces the misjudgment rate of a single model. The results show that the integrated weight of the improved machine learning algorithm based on ROS improves the accuracy of rock mass quality evaluation model, and provides a new idea for rock mass quality evaluation. |
| Close |
|
|
|