Expand description
Optimal transport kernel approximation methods
This module implements kernel approximation methods based on optimal transport theory, including Wasserstein kernels, Sinkhorn divergences, and earth mover’s distance approximations.
Structs§
- EMDKernel
Sampler - Earth Mover’s Distance (EMD) kernel approximation
- FittedEMD
Sampler - Fitted EMD kernel sampler
- Fitted
Gromov Wasserstein Sampler - Fitted Gromov-Wasserstein sampler
- Fitted
Wasserstein Sampler - Fitted Wasserstein kernel sampler
- Gromov
Wasserstein Sampler - Gromov-Wasserstein kernel approximation
- Wasserstein
Kernel Sampler - Wasserstein kernel approximation using random features
Enums§
- GWLoss
Function - Loss functions for Gromov-Wasserstein GWLossFunction
- Ground
Metric - Ground metric options for optimal transport GroundMetric
- Transport
Method - Transport plan approximation methods TransportMethod