Unsupervised learning for blind source separation: an information-theoretic approach.
Dragan ObradovicGustavo DecoPublished in: ICASSP (1997)
Keyphrases
- blind source separation
- unsupervised learning
- component analysis
- independent component analysis
- non stationary
- supervised learning
- adaptive particle swarm optimization
- data clustering
- source separation
- natural gradient
- negative matrix factorization
- blind separation
- mixing matrix
- speech signal
- eigenvalue decomposition
- semi supervised
- sparse representation
- dimensionality reduction
- machine learning
- cluster analysis
- object recognition
- independent components
- text classification
- algebraic equations
- factor analysis
- model selection
- reinforcement learning