Multiview Boosting by Controlling the ...
Document type :
Article dans une revue scientifique: Article original
Title :
Multiview Boosting by Controlling the Diversity and the Accuracy of View-specific Voters
Author(s) :
Goyal, Anil [Auteur]
Laboratoire Hubert Curien [LabHC]
Analyse de données, Modélisation et Apprentissage automatique [Grenoble] [AMA ]
Morvant, Emilie [Auteur]
Laboratoire Hubert Curien [LabHC]
Germain, Pascal [Auteur]
MOdel for Data Analysis and Learning [MODAL]
Amini, Massih-Reza [Auteur]
Analyse de données, Modélisation et Apprentissage automatique [Grenoble] [AMA ]
Laboratoire Hubert Curien [LabHC]
Analyse de données, Modélisation et Apprentissage automatique [Grenoble] [AMA ]
Morvant, Emilie [Auteur]
Laboratoire Hubert Curien [LabHC]
Germain, Pascal [Auteur]
MOdel for Data Analysis and Learning [MODAL]
Amini, Massih-Reza [Auteur]
Analyse de données, Modélisation et Apprentissage automatique [Grenoble] [AMA ]
Journal title :
Neurocomputing
Pages :
81-92
Publisher :
Elsevier
Publication date :
2019-09-17
ISSN :
0925-2312
English keyword(s) :
PAC-Bayes
Multiview Learning
Boosting
Multiview Learning
Boosting
HAL domain(s) :
Statistiques [stat]/Machine Learning [stat.ML]
English abstract : [en]
In this paper we propose a boosting based multiview learning algorithm, referred to as PB-MVBoost, which iteratively learns i) weights over view-specific voters capturing view-specific information; and ii) weights over ...
Show more >In this paper we propose a boosting based multiview learning algorithm, referred to as PB-MVBoost, which iteratively learns i) weights over view-specific voters capturing view-specific information; and ii) weights over views by optimizing a PAC-Bayes multiview C-Bound that takes into account the accuracy of view-specific classifiers and the diversity between the views. We derive a generalization bound for this strategy following the PAC-Bayes theory which is a suitable tool to deal with models expressed as weighted combination over a set of voters. Different experiments on three publicly available datasets show the efficiency of the proposed approach with respect to state-of-art models.Show less >
Show more >In this paper we propose a boosting based multiview learning algorithm, referred to as PB-MVBoost, which iteratively learns i) weights over view-specific voters capturing view-specific information; and ii) weights over views by optimizing a PAC-Bayes multiview C-Bound that takes into account the accuracy of view-specific classifiers and the diversity between the views. We derive a generalization bound for this strategy following the PAC-Bayes theory which is a suitable tool to deal with models expressed as weighted combination over a set of voters. Different experiments on three publicly available datasets show the efficiency of the proposed approach with respect to state-of-art models.Show less >
Language :
Anglais
Peer reviewed article :
Oui
Audience :
Internationale
Popular science :
Non
ANR Project :
Collections :
Source :
Files
- document
- Open access
- Access the document
- PBMVBoost_Paper.pdf
- Open access
- Access the document
- hierarchy.pdf
- Open access
- Access the document
- histograms.png
- Open access
- Access the document
- MNIST_1_500_iterations.png
- Open access
- Access the document
- MNIST_1_acc.png
- Open access
- Access the document
- MNIST_1_F1.png
- Open access
- Access the document
- MNIST_2_500_iterations.png
- Open access
- Access the document
- MNIST_2_acc.png
- Open access
- Access the document
- MNIST_2_F1.png
- Open access
- Access the document
- Reuters_500_iterations.png
- Open access
- Access the document
- Reuters_acc.png
- Open access
- Access the document
- Reuters_F1.png
- Open access
- Access the document
- 1808.05784
- Open access
- Access the document