Geometric Feature Learning for Functional Data Valued on the Symmetric Positive Definite Manifold
Abstract
We here develop a functional neural network, termed MatFAE, for learning trajectories on the Riemannian manifold of symmetric positive definite (SPD) matrices. MatFAE features intrinsic layers that map manifold-valued functions to Euclidean vector-valued functions, followed by a functional layer that projects them into a finite-dimensional Euclidean space. Unlike most neural networks for discrete-time sequences, MatFAE treats each sequence as a continuous function and can therefore encode trajectory dynamics (e.g., first-order derivatives) in its latent representations. Additionally, the morphology of the functional weights in the functional layer offers interpretability by revealing the regions of the input functional data that contribute most to the latent representations. We justify the design principles and properties of each intrinsic layer and detail how matrix factorization is handled during backpropagation. We apply MatFAE to a range of fMRI datasets, demonstrating its ability to efficiently learn informative representations from high-dimensional SPD trajectories and its practical value for real-world neuroimaging analysis.
Related
- Convergence analysis of TriCG and TriMR for symmetric quasi-definite linear systems
- Distributionally Faithful Imputation via Positive Semi-Definite Kernel Density Estimation
- A positive-definite Fourier transform, with applications to fractional Brownian motions and Bayesian probabilistic numerics
- Geodesics and intrinsic statistics of isolated connections in Hermitian positive-definite matrices
- Structured Positive-Definite Optimal Control Synthesis of Closed-Loop Recommendation Systems over Social Networks
BibTeX
@misc{singh2026geometric,
title = {{Geometric Feature Learning for Functional Data Valued on the Symmetric Positive Definite Manifold}},
author = {Samuel V. Singh and Mimi Zhang},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.30487},
eprint = {2609.30487},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.30487},
note = {CC BY-NC-SA 4.0},
}
RIS
TY - UNPB
TI - Geometric Feature Learning for Functional Data Valued on the Symmetric Positive Definite Manifold
AU - Samuel V. Singh
AU - Mimi Zhang
PY - 2026
DA - 2026/09/24
PB - arXiv
DO - 10.48550/arxiv.2609.30487
UR - https://arxiv.org/abs/2609.30487
AB - We here develop a functional neural network, termed MatFAE, for learning trajectories on the Riemannian manifold of symmetric positive definite (SPD) matrices. MatFAE features intrinsic layers that map manifold-valued functions to Euclidean vector-valued functions, followed by a functional layer that projects them into a finite-dimensional Euclidean space. Unlike most neural networks for discrete-time sequences, MatFAE treats each sequence as a continuous function and can therefore encode trajectory dynamics (e.g., first-order derivatives) in its latent representations. Additionally, the morphology of the functional weights in the functional layer offers interpretability by revealing the regions of the input functional data that contribute most to the latent representations. We justify the design principles and properties of each intrinsic layer and detail how matrix factorization is handled during backpropagation. We apply MatFAE to a range of fMRI datasets, demonstrating its ability to efficiently learn informative representations from high-dimensional SPD trajectories and its practical value for real-world neuroimaging analysis.
ER -