Semi-supervised rotation forest based on ensemble margin theory for the classification of hyperspectral image with limited training data.
Wei FengYinghui QuanGabriel DauphinQiang LiLianru GaoWenjiang HuangJunshi XiaWentao ZhuMengdao XingPublished in: Inf. Sci. (2021)
Keyphrases
- training set
- training data
- hyperspectral images
- supervised learning
- semi supervised
- decision trees
- classification accuracy
- ensemble methods
- rotation forest
- random forest
- ensemble classifier
- hyperspectral data
- base classifiers
- class labels
- labeled data
- unlabeled data
- classification models
- semi supervised learning
- training samples
- support vector machine
- active learning
- hyperspectral
- hyperspectral imagery
- classifier ensemble
- benchmark datasets
- classification algorithm
- feature selection
- generalization error
- learning algorithm
- multispectral
- machine learning methods
- training examples
- svm classifier
- individual classifiers
- ensemble learning
- remote sensing
- generalization ability
- support vector machine svm
- prediction accuracy
- target detection
- feature set
- data sets
- test data
- image classification
- knn
- pairwise
- feature space