Learning algorithms utilizing quasi-geodesic flows on the Stiefel manifold.
Yasunori NishimoriShotaro AkahoPublished in: Neurocomputing (2005)
Keyphrases
- euclidean space
- geodesic distance
- learning algorithm
- riemannian metric
- riemannian manifolds
- shape analysis
- active learning
- euclidean distance
- supervised learning
- shape space
- machine learning algorithms
- manifold valued
- machine learning
- learning problems
- vector space
- graph laplacian
- low dimensional
- shortest path
- manifold learning
- metric space
- learning tasks
- fisher information
- reinforcement learning
- training data
- back propagation
- distance transform
- parameter space
- data points
- learning scheme
- subspace learning
- feature space
- learning rate
- heat kernel
- geodesic paths
- tangent space