The decomposed normalized maximum likelihood code-length criterion for selecting hierarchical latent variable models.
Kenji YamanishiTianyi WuShinya SugawaraMakoto OkadaPublished in: Data Min. Knowl. Discov. (2019)
Keyphrases
- normalized maximum likelihood
- latent variable models
- code length
- minimum description length
- latent variables
- change point detection
- gray code
- mdl principle
- spatially adaptive
- hidden markov models
- information theoretic
- non stationary
- learning problems
- latent space
- channel coding
- feature selection
- principal component analysis
- probabilistic model
- feature extraction
- image segmentation