Structure Learning via Hadamard Product of Correlation and Partial Correlation Matrices

Karina Ashurbekova; Sophie Achard; Florence Forbes

doi:10.23919/EUSIPCO.2019.8902948

Communication dans un congrès

Structure Learning via Hadamard Product of Correlation and Partial Correlation Matrices

Karina Ashurbekova ^{1, 2} Sophie Achard ² Florence Forbes ¹

1 MISTIS [2016-2019] - Modelling and Inference of Complex and Structured Stochastic Systems [2016-2019]

Inria Grenoble - Rhône-Alpes, Grenoble INP [2007-2019] - Institut polytechnique de Grenoble - Grenoble Institute of Technology [2007-2019], LJK - Laboratoire Jean Kuntzmann

2 GIPSA-VIBS - GIPSA - Vision and Brain Signal Processing

GIPSA-DIS - Département Images et Signal

Abstract : Structure learning is an active topic nowadays in different application areas, i.e. genetics, neuroscience. Classical conditional independences or marginal independences may not be sufficient to express complex relationships. This paper is introducing a new structure learning procedure where an edge in the graph corresponds to a non zero value of both correlation and partial correlation. Based on this new paradigm, we define an estimator and derive its theoretical properties. The asymptotic convergence of the proposed graph estimator and its rate are derived. Illustrations on a synthetic example and application to brain connectivity are displayed.

Keywords : brain connectivity conditional independence structure learning marginal independence soft-thresholding Hadamard product

Type de document :

Communication dans un congrès

Domaine :

Statistiques [stat] / Méthodologie [stat.ME]

Statistiques [stat] / Applications [stat.AP]

Liste complète des métadonnées

Littérature citée [23 références] Voir Masquer Télécharger

https://hal.univ-grenoble-alpes.fr/hal-02290847
Contributeur : Sophie Achard <>
Soumis le : jeudi 19 septembre 2019 - 09:00:03
Dernière modification le : mercredi 14 octobre 2020 - 03:51:38
Archivage à long terme le : : dimanche 9 février 2020 - 00:59:57