Harnessing Orthogonality to Train Low-Rank Neural Networks.
Daniel CoquelinKatharina FlügelMarie WeielNicholas KieferCharlotte DebusAchim StreitMarkus GötzPublished in: CoRR (2024)
Keyphrases
- low rank
- neural network
- convex optimization
- linear combination
- missing data
- singular value decomposition
- matrix factorization
- semi supervised
- matrix completion
- matrix decomposition
- low rank matrix
- rank minimization
- high dimensional data
- nonnegative matrix factorization
- kernel matrix
- pattern recognition
- high order
- trace norm
- minimization problems
- low rank matrices
- affinity matrix
- low rank representation
- robust principal component analysis
- dimensionality reduction
- data matrix
- singular values
- small number
- data sets
- decision trees
- non rigid structure from motion
- low rank and sparse
- regularized regression