A Lagrangian-based score for assessing the quality of pairwise constraints in semi-supervised clustering.
Rodrigo RandelDaniel AloiseSimon J. BlanchardAlain HertzPublished in: Data Min. Knowl. Discov. (2021)
Keyphrases
- semi supervised clustering
- pairwise constraints
- semi supervised
- pairwise
- semi supervised classification
- document clustering
- unsupervised clustering
- background knowledge
- data representation
- spectral clustering
- loss function
- unlabeled data
- decision boundary
- metric learning
- clustering algorithm
- labeled data
- semi supervised learning
- distance metric
- nonnegative matrix factorization
- k means
- natural language processing
- training data
- hidden markov random fields
- multi class
- active learning
- machine learning
- constrained clustering