Cramer-Rao lower bounds for low-rank decomposition of multidimensional arrays.
Xiangqian LiuNikos D. SidiropoulosPublished in: IEEE Trans. Signal Process. (2001)
Keyphrases
- low rank
- cramer rao
- lower bound
- multidimensional arrays
- fisher information
- error bounds
- noise level
- linear combination
- upper bound
- convex optimization
- missing data
- matrix factorization
- singular value decomposition
- image registration
- semi supervised
- high order
- kernel matrix
- high dimensional data
- objective function
- database management systems
- worst case
- data structure
- np hard
- optimal solution
- multidimensional data
- data cube
- positive definite
- pairwise
- probability density function
- noisy images
- maximum likelihood
- active learning