Toeplitz Hermitian Positive Definite Matrix Machine Learning Based on Fisher Metric.
Yann CabanesFrédéric BarbarescoMarc ArnaudonJérémie BigotPublished in: GSI (2019)
Keyphrases
- positive definite
- fisher information
- machine learning
- covariance matrix
- kernel function
- geometric structure
- kernel matrix
- diffusion tensor
- reproducing kernel hilbert space
- learning algorithm
- least squares
- machine learning algorithms
- kernel methods
- support vector machine
- distance measure
- semi parametric
- metric learning
- decision trees
- symmetric positive definite
- learning tasks
- kernel machines
- metric space
- sufficient conditions
- active learning
- symmetric matrices
- special case