PCA-based SVM classification for simulated ice floes in front of sluice gates
The entrainment and accumulation of ice floes in front of sluice gates are closely related to the water transport efficiency and safe operation of the channel during the ice period. A flume study is carried out for a sluice gate with free outflow. An integrated principal component analysis and suppo...
| Published in: | Polar Science |
|---|---|
| Format: | Article in Journal/Newspaper |
| Language: | English |
| Published: |
2022
|
| Subjects: | |
| Online Access: | https://nipr.repo.nii.ac.jp/?action=repository_uri&item_id=16989 http://id.nii.ac.jp/1291/00016858/ |
| Summary: | The entrainment and accumulation of ice floes in front of sluice gates are closely related to the water transport efficiency and safe operation of the channel during the ice period. A flume study is carried out for a sluice gate with free outflow. An integrated principal component analysis and support vector machine (PCA-SVM) model for simulated ice floes classification is proposed. Based on the mechanism of ice floe accumulation, ten input characteristics of the model are selected. The first principal component, with a contribution rate of 71.76%, and the second principal component, with a contribution rate of 15.64%, are extracted as the inputs of the SVM model. The 5-fold cross-validation method is used to examine the model. The training results show that the Gaussian radial basis function (RBF) is the optimal kernel function. The performance of the developed model is measured by a confusion matrix and receiver operating characteristic (ROC) analysis. The results show that the established PCA-SVM model improves upon the Bernoulli naive Bayes (Bernoulli NB) and K-nearest neighbor (KNN) models, increasing the area under the ROC curve (AUC) values by 11% and 5%, the accuracy (Acc) values by 16% and 17%, and the F1 values by 17% and 2%, respectively. |
|---|