Paper Outline¶
Introduction¶
What is Similarity?
- Invariances
- Orthogonal
- Scaling
- Multivariate
- Non-Linear
Background¶
- Linear Methods
- RV Coefficient
- Kernel methods
- hsic
- cka
- information metrics
- entropy
- mutual information
- normalize mutual information
Methods¶
Gaussianization Flows¶
- Density Flows
- Generalized Form Gaussianization
- RBIG in particular
Information Measures w. RBIG¶
- Loss Function w. Total Correlation
- H, trivial to calculate
- MI, 3 RBIGs
Experiments Background¶
Data¶
- ESDC
- Variables
- GPP
- LST
- SM
- Density Cubes
Algorithm Details¶
- RV Coefficient
- cKA Method
- RBIG Details