Some improvements of the spectral learning approach for probabilistic grammatical inference - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2014

Some improvements of the spectral learning approach for probabilistic grammatical inference

Résumé

Spectral methods propose new and elegant solutions in probabilistic grammatical inference. We propose two ways to improve them. We show how a linear representation, or equivalently a weighted automata, output by the spectral learning algorithm can be taken as an initial point for the Baum Welch algorithm, in order to increase the likelihood of the observation data. Secondly, we show how the inference problem can naturally be expressed in the framework of Structured Low-Rank Approximation. Both ideas are tested on a benchmark extracted from the PAutomaC challenge.
Fichier principal
Vignette du fichier
gybels14a.pdf (368.35 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01075979 , version 1 (20-10-2014)

Identifiants

  • HAL Id : hal-01075979 , version 1

Citer

Mattias Gybels, Francois Denis, Amaury Habrard. Some improvements of the spectral learning approach for probabilistic grammatical inference. Proceedings of the 12th International Conference on Grammatical Inference (ICGI), Sep 2014, Kyoto, Japan. pp.64-78. ⟨hal-01075979⟩
105 Consultations
50 Téléchargements

Partager

Gmail Facebook X LinkedIn More