Sacrificing Overall Classification Quality to Improve Classification Accuracy of Well-Sought Classes.
Kevin Michael AmaralPing ChenWei DingRajani S. SadasivamPublished in: ICDM Workshops (2016)
Keyphrases
- class labels
- classification accuracy
- machine learning methods
- automatic classification
- classification systems
- object classification
- support vector machine
- pattern classification
- classification scheme
- image classification
- feature vectors
- pattern recognition
- multiple classes
- decision trees
- discriminant functions
- high quality
- classification algorithm
- machine learning
- classification rules
- decision rules
- classification method
- svm classifier
- training samples
- rare class
- neural network
- feature extraction
- binary classifiers
- support vector
- image quality
- class imbalance
- support vector machine svm
- roc curve
- quality assessment
- quality measures
- benchmark data sets
- data quality