for Information-Theoretic Classification Accuracy: A Criterion That Guides Data-Driven Combination of Ambiguous Outcome Labels in Multiclass Classification.
Chihao ZhangShihua ZhangJingyi Jessica LiPublished in: J. Comput. Biol. (2023)
Keyphrases
- information theoretic
- multiclass classification
- classification accuracy
- minimum description length
- multi class
- mutual information
- information theory
- feature selection
- relative entropy
- binary classification
- training data
- training set
- support vector
- multi label
- binary classifiers
- information theoretic measures
- pairwise
- selection criterion
- class labels
- support vector machine
- cost sensitive
- training examples
- similarity measure
- machine learning
- image classification
- text classification
- model selection
- nearest neighbor
- active learning
- pattern recognition
- feature extraction
- decision trees