Unsupervised Linear Discriminant Analysis for Supporting DPGMM Clustering in the Zero Resource Scenario.
Michael HeckSakriani SaktiSatoshi NakamuraPublished in: SLTU (2016)
Keyphrases
- linear discriminant analysis
- dealing with high dimensional data
- unsupervised feature selection
- discriminant analysis
- high dimensional data
- dimensionality reduction
- unsupervised learning
- principal components analysis
- discriminant projection
- face recognition
- dimension reduction
- principal component analysis
- discriminant features
- class separability
- fisher criterion
- small sample size
- clustering algorithm
- feature extraction
- subspace methods
- linear discriminant
- support vector machine svm
- feature space
- null space
- clustering method
- support vector
- cluster analysis
- data points
- high dimensionality
- discriminative information
- semi supervised
- data clustering
- generalized discriminant analysis
- nearest neighbor
- feature selection
- qr decomposition
- pattern recognition
- high dimensional
- machine learning
- scatter matrix
- scatter matrices
- multi class
- supervised learning
- low dimensional
- dimensionality reduction methods
- input data
- lower dimensional
- distance metric
- neural network