Probabilistic Semi-Nonnegative Matrix Factorization via Maximum-Likelihood and Variational Inference.
Junbin LiuMingjie ShaoWing-Kin MaPublished in: EUSIPCO (2023)
Keyphrases
- nonnegative matrix factorization
- variational inference
- maximum likelihood
- probabilistic model
- exponential family
- mixture model
- bayesian inference
- posterior distribution
- markov networks
- expectation maximization
- probabilistic graphical models
- data representation
- em algorithm
- parameter estimation
- hyperparameters
- maximum a posteriori
- negative matrix factorization
- matrix factorization
- generative model
- latent dirichlet allocation
- topic models
- factor graphs
- least squares
- bayesian networks
- gaussian process
- variational methods
- spectral clustering
- principal component analysis
- gaussian distribution
- posterior probability
- bayesian framework
- parameter learning
- closed form
- latent variables
- original data
- document clustering
- approximate inference
- language model
- probability distribution
- prior knowledge