Recurrent Neural Network Compression Based on Low-Rank Tensor Representation.
Andros TjandraSakriani SaktiSatoshi NakamuraPublished in: IEICE Trans. Inf. Syst. (2020)
Keyphrases
- recurrent neural networks
- low rank
- linear combination
- missing data
- matrix factorization
- convex optimization
- low rank matrix
- rank minimization
- matrix completion
- singular value decomposition
- neural network
- high order
- complex valued
- semi supervised
- feed forward
- recurrent networks
- high dimensional data
- image compression
- matrix decomposition
- low rank matrices
- singular values
- artificial neural networks
- robust principal component analysis
- trace norm
- higher order
- regularized regression
- learning algorithm
- similarity measure
- high dimensional
- active learning
- small number
- nearest neighbor
- real valued
- data sets