Geometric statistics with subspace structure preservation for SPD matrices

Geometric statistics with subspace structure preservation for SPD matrices

We present a geometric framework for the processing of SPD-valued data that preserves subspace structures and is based on the efficient computation of extreme generalized eigenvalues. This is achieved through the use of the Thompson geometry of the semidefinite cone. We explore a particular geodesic...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Mostajeran, Cyrus, Da Costa, Nathaël, Van Goffrier, Graham, Sepulchre, Rodolphe
Format: Artikel
Sprache:eng
Schlagworte:
Computer Science - Learning
Computer Science - Numerical Analysis
Mathematics - Differential Geometry
Mathematics - Numerical Analysis
Statistics - Computation
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:We present a geometric framework for the processing of SPD-valued data that preserves subspace structures and is based on the efficient computation of extreme generalized eigenvalues. This is achieved through the use of the Thompson geometry of the semidefinite cone. We explore a particular geodesic space structure in detail and establish several properties associated with it. Finally, we review a novel inductive mean of SPD matrices based on this geometry.
DOI:10.48550/arxiv.2407.03382