Gradient Descent Happens in a Tiny Subspace.
Guy Gur-AriDaniel A. RobertsEthan DyerPublished in: CoRR (2018)
Keyphrases
- cost function
- low dimensional
- clustering high dimensional data
- subspace clustering
- loss function
- linear subspace
- subspace learning
- high dimensional
- dimensionality reduction
- objective function
- principal component analysis
- lower dimensional
- kernel based nonlinear
- principal components
- high dimensional data
- feature space
- feature extraction
- image processing
- principal components analysis
- computer vision
- database