Real-Time Tear Film Classification Through Cost-Based Feature Selection.
Verónica Bolón-CanedoBeatriz RemeseiroNoelia Sánchez-MaroñoAmparo Alonso-BetanzosPublished in: Trans. Comput. Collect. Intell. (2015)
Keyphrases
- feature selection
- real time
- classification accuracy
- support vector
- pattern recognition
- text classification
- machine learning
- classification models
- feature selection algorithms
- model selection
- feature set
- high dimensionality
- feature space
- cost sensitive
- class imbalance
- feature extraction
- mutual information
- feature vectors
- unsupervised learning
- accurate classification
- multi class
- informative features
- support vector machine svm
- cost sensitive learning
- irrelevant features
- method for feature selection
- support vector machine
- discriminative features
- feature subset
- decision trees
- total cost
- supervised learning
- select relevant features
- conditional mutual information
- feature selection and classification
- classification performances
- small samples
- class separability
- preprocessing
- web page classification
- image classification
- class labels
- small sample
- feature ranking
- classification algorithm
- ensemble classifier
- selected features
- classification rate
- information gain