Fast online low-rank tensor subspace tracking by CP decomposition using recursive least squares from incomplete observations.
Hiroyuki KasaiPublished in: Neurocomputing (2019)
Keyphrases
- low rank
- tensor decomposition
- recursive least squares
- missing data
- rank minimization
- trace norm
- high dimensional data
- high order
- kalman filtering
- linear combination
- singular value decomposition
- convex optimization
- matrix factorization
- dimensionality reduction
- regularized regression
- matrix completion
- kalman filter
- principal component analysis
- particle filter
- semi supervised
- adaptive filtering
- low dimensional
- neuro fuzzy
- visual tracking
- subspace learning
- image processing
- higher order
- structure from motion
- neural network
- high dimensional
- optical flow
- input data
- nearest neighbor
- autoregressive
- motion estimation
- data points
- object tracking
- basis functions
- pairwise
- pattern recognition
- reinforcement learning
- feature extraction
- appearance model