Supervised nonnegative matrix factorization with Dual-Itakura-Saito and Kullback-Leibler divergences for music transcription.
Hideaki KagamiMasahiro YukawaPublished in: EUSIPCO (2016)
Keyphrases
- bregman divergences
- kullback leibler
- nonnegative matrix factorization
- kl divergence
- relative entropy
- kullback leibler divergence
- distance measure
- negative matrix factorization
- matrix factorization
- cross entropy
- semi supervised
- mahalanobis distance
- data representation
- information theoretic
- supervised learning
- principal component analysis
- least squares
- spectral clustering
- unsupervised learning
- document clustering
- machine learning
- gaussian distribution
- exponential family
- feature selection
- information theory
- posterior distribution
- gaussian mixture
- graphical models