A Semi-Definite Programming approach to low dimensional embedding for unsupervised clustering.
Stéphane ChrétienClément DombryAdrien FaivrePublished in: CoRR (2016)
Keyphrases
- unsupervised clustering
- semi definite programming
- low dimensional
- pairwise constraints
- semi supervised
- high dimensional
- supervised classification
- pairwise
- dimensionality reduction
- spectral clustering
- kernel matrix
- data points
- metric learning
- input space
- feature space
- clustering method
- high dimensional data
- euclidean space
- k means
- principal component analysis
- labeled data
- loss function
- unlabeled data
- document clustering
- semi supervised learning
- data representation
- clustering algorithm
- distance metric
- fuzzy c means
- image segmentation
- pattern recognition
- decision trees
- maximum likelihood
- feature selection
- supervised learning
- decision boundary