Alternating linear scheme in a Bayesian framework for low-rank tensor approximation.
Clara MenzenManon KokKim BatselierPublished in: CoRR (2020)
Keyphrases
- bayesian framework
- low rank
- frobenius norm
- low rank matrix
- trace norm
- high order
- tensor decomposition
- matrix factorization
- convex optimization
- missing data
- prior knowledge
- matrix completion
- generative model
- approximation methods
- linear combination
- singular value decomposition
- hyperparameters
- semi supervised
- gaussian process
- low rank approximation
- kernel matrix
- expectation maximization
- rank minimization
- probability density function
- maximum a posteriori
- high dimensional data
- higher order
- closed form
- computer vision
- decision trees
- feature extraction
- probabilistic model