Generalized Learning Vector Quantization With Log-Euclidean Metric Learning on Symmetric Positive-Definite Manifold.
Fengzhen TangPeter TinoHaibin YuPublished in: IEEE Trans. Cybern. (2023)
Keyphrases
- spd matrices
- metric learning
- symmetric positive definite
- kernel matrix
- positive definite
- log euclidean
- distance metric
- riemannian manifolds
- distribution function
- pairwise
- lie group
- dimensionality reduction
- semi supervised
- linear transformation
- multi task
- learning tasks
- diffusion tensor
- reproducing kernel hilbert space
- feature space
- distance function
- riemannian metric
- affine invariant
- euclidean space
- vector space
- geometric structure
- covariance matrices
- similarity measure
- kernel pca
- finite dimensional
- tensor field
- neural network
- reinforcement learning
- feature selection
- machine learning