Accelerate convolutional neural networks for binary classification via cascading cost-sensitive feature.
Junbiao PangHuihuang LinLi SuChunjie ZhangWeigang ZhangLijuan DuanQingming HuangBaocai YinPublished in: ICIP (2016)
Keyphrases
- binary classification
- cost sensitive
- convolutional neural networks
- multi class
- misclassification costs
- cost sensitive learning
- multi class classification
- cost sensitive classification
- class distribution
- class imbalance
- active learning
- naive bayes
- boosting algorithms
- binary classifiers
- active learning strategies
- support vector machine
- feature vectors
- learning problems
- feature set
- image features
- base classifiers
- ensemble methods
- image classification
- decision trees