Positive-Unlabeled Classification under Class Prior Shift and Asymmetric Error.
Nontawat CharoenphakdeeMasashi SugiyamaPublished in: SDM (2019)
Keyphrases
- class labels
- positive and unlabeled examples
- supervised learning
- feature extraction
- unsupervised learning
- training data
- training set
- classification accuracy
- multi class classification
- automatic classification
- classification algorithm
- prior probabilities
- image classification
- multiclass classification
- positive examples
- positive and negative
- error rate
- feature selection
- machine learning
- prior knowledge
- text classification
- bayesian methods
- pattern recognition
- decision trees
- class specific
- classification rules
- feature vectors
- class membership
- multi class
- labeled data
- support vector machine
- binary classification problems
- class conditional
- positive data
- intra class
- neural network
- decision boundary
- preprocessing
- document classification
- classification models
- pattern classification
- feature space
- cross validation
- benchmark datasets
- data sets
- training examples
- support vector machine svm