PAC-Bayesian bounds for learning LTI-ss systems with input from empirical loss

Eringis, Deividas; Leth, John; Tan, Zheng-Hua; Wisniewski, Rafael; Petreczky, Mihaly

Type de document :

Autre communication scientifique (congrès sans actes - poster - séminaire...): Communication dans un congrès avec actes

DOI :

10.48550/arXiv.2303.16816

Titre :

PAC-Bayesian bounds for learning LTI-ss systems with input from empirical loss

Auteur(s) :

Eringis, Deividas [Auteur]
Department of Electronic Systems - Aalborg University
Leth, John [Auteur]
Department of Electronic Systems - Aalborg University
Tan, Zheng-Hua [Auteur]
Department of Computer Science [Aalborg]
Wisniewski, Rafael [Auteur]
Department of Electronic Systems - Aalborg University
Petreczky, Mihaly [Auteur]

Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]

Titre de la manifestation scientifique :

Workshop Frontiers4LCD ICML 2023

Ville :

Honolulu

Pays :

Etats-Unis d'Amérique

Date de début de la manifestation scientifique :

2023-07

Date de publication :

2023-07

Mot(s)-clé(s) en anglais :

Machine Learning (stat.ML)
Machine Learning (cs.LG)
FOS: Computer and information sciences

Discipline(s) HAL :

Informatique [cs]/Intelligence artificielle [cs.AI]
Sciences de l'ingénieur [physics]/Automatique / Robotique

Résumé en anglais : [en]

In this paper we derive a Probably Approxilmately Correct(PAC)-Bayesian error bound for linear time-invariant (LTI) stochastic dynamical systems with inputs. Such bounds are widespread in machine learning, and they are ...
Lire la suite >In this paper we derive a Probably Approxilmately Correct(PAC)-Bayesian error bound for linear time-invariant (LTI) stochastic dynamical systems with inputs. Such bounds are widespread in machine learning, and they are useful for characterizing the predictive power of models learned from finitely many data points. In particular, with the bound derived in this paper relates future average prediction errors with the prediction error generated by the model on the data used for learning. In turn, this allows us to provide finite-sample error bounds for a wide class of learning/system identification algorithms. Furthermore, as LTI systems are a sub-class of recurrent neural networks (RNNs), these error bounds could be a first step towards PAC-Bayesian bounds for RNNs.Lire moins >

Langue :

Anglais

Comité de lecture :

Oui

Audience :

Internationale

Vulgarisation :

Non

Commentaire :

arXiv admin note: text overlap with arXiv:2212.14838

Collections :

Centre de Recherche en Informatique, Signal et Automatique de Lille (CRIStAL) - UMR 9189

Source :

Harvested from HAL

Fichiers

document
Accès libre
Accéder au document

2303.16816%281%29.pdf
Accès libre
Accéder au document

2303.16816
Accès libre
Accéder au document

PAC-Bayesian bounds for learning LTI-ss ... BibTeX CSV Excel RIS

Fichiers

PAC-Bayesian bounds for learning LTI-ss ...

BibTeX

CSV

Excel

RIS