Word Embedding Based on Low-Rank Doubly Stochastic Matrix Decomposition.
Denis SedovZhirong YangPublished in: ICONIP (3) (2018)
Keyphrases
- matrix decomposition
- low rank
- low rank matrix
- convex optimization
- linear combination
- missing data
- matrix factorization
- singular value decomposition
- kernel matrix
- high order
- nonnegative matrix factorization
- semi supervised
- high dimensional data
- data matrix
- singular values
- vector space
- low rank approximation
- pairwise
- spectral clustering