Theory#
Mathematical foundations and reference material for SPD Learn.
Start here
Background
- Data representations (EEG/fMRI covariances)
- Minimal geometry & metric choices
- SPDNet pipeline overview
Core math
Geometric Concepts
- SPD manifold & Riemannian metrics
- Exp/Log maps, parallel transport
- Layer operation visualizations
Practical
Numerical Stability
- Dtype-aware thresholds & clamping
- Configuration for stability
- Troubleshooting NaN & convergence
Reference
Notation
- Manifold symbols (𝒮⁺, Sym, T_P)
- Distance & metric conventions
- Layer operation symbols
Reference
Glossary
- SPD matrices & Riemannian terms
- Layer & module definitions
- Model architecture names
Reference
References
- Interactive literature map
- Model family legend
- Full bibliography & BibTeX
See also
User Guide – Getting started with SPD Learn
API Reference – Complete API reference
Applied Examples – Practical examples