Multiview Boosting With Information Propagation for Classification

Jing Peng, Alex J. Aved, Guna Seetharaman, Kannappan Palaniappan

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Multiview learning has shown promising potential in many applications. However, most techniques are focused on either view consistency, or view diversity. In this paper, we introduce a novel multiview boosting algorithm, called Boost.SH, that computes weak classifiers independently of each view but uses a shared weight distribution to propagate information among the multiple views to ensure consistency. To encourage diversity, we introduce randomized Boost.SH and show its convergence to the greedy Boost.SH solution in the sense of minimizing regret using the framework of adversarial multiarmed bandits. We also introduce a variant of Boost.SH that combines decisions from multiple experts for recommending views for classification. We propose an expert strategy for multiview learning based on inverse variance, which explores both consistency and diversity. Experiments on biometric recognition, document categorization, multilingual text, and yeast genomic multiview data sets demonstrate the advantage of Boost.SH (85%) compared with other boosting algorithms like AdaBoost (82%) using concatenated views and substantially better than a multiview kernel learning algorithm (74%).

Original languageEnglish
Pages (from-to)657-669
Number of pages13
JournalIEEE Transactions on Neural Networks and Learning Systems
Volume29
Issue number3
DOIs
StatePublished - 1 Mar 2018

Fingerprint

Adaptive boosting
Biometrics
Yeast
Learning algorithms
Classifiers
Experiments

Keywords

  • Biometrics
  • boosting
  • classification
  • convergence
  • data fusion
  • multiarmed bandits
  • multiview learning

Cite this

Peng, Jing ; Aved, Alex J. ; Seetharaman, Guna ; Palaniappan, Kannappan. / Multiview Boosting With Information Propagation for Classification. In: IEEE Transactions on Neural Networks and Learning Systems. 2018 ; Vol. 29, No. 3. pp. 657-669.
@article{64d33205bef141c48e1085ff3fffaabc,
title = "Multiview Boosting With Information Propagation for Classification",
abstract = "Multiview learning has shown promising potential in many applications. However, most techniques are focused on either view consistency, or view diversity. In this paper, we introduce a novel multiview boosting algorithm, called Boost.SH, that computes weak classifiers independently of each view but uses a shared weight distribution to propagate information among the multiple views to ensure consistency. To encourage diversity, we introduce randomized Boost.SH and show its convergence to the greedy Boost.SH solution in the sense of minimizing regret using the framework of adversarial multiarmed bandits. We also introduce a variant of Boost.SH that combines decisions from multiple experts for recommending views for classification. We propose an expert strategy for multiview learning based on inverse variance, which explores both consistency and diversity. Experiments on biometric recognition, document categorization, multilingual text, and yeast genomic multiview data sets demonstrate the advantage of Boost.SH (85{\%}) compared with other boosting algorithms like AdaBoost (82{\%}) using concatenated views and substantially better than a multiview kernel learning algorithm (74{\%}).",
keywords = "Biometrics, boosting, classification, convergence, data fusion, multiarmed bandits, multiview learning",
author = "Jing Peng and Aved, {Alex J.} and Guna Seetharaman and Kannappan Palaniappan",
year = "2018",
month = "3",
day = "1",
doi = "10.1109/TNNLS.2016.2637881",
language = "English",
volume = "29",
pages = "657--669",
journal = "IEEE Transactions on Neural Networks and Learning Systems",
issn = "2162-237X",
publisher = "IEEE Computational Intelligence Society",
number = "3",

}

Multiview Boosting With Information Propagation for Classification. / Peng, Jing; Aved, Alex J.; Seetharaman, Guna; Palaniappan, Kannappan.

In: IEEE Transactions on Neural Networks and Learning Systems, Vol. 29, No. 3, 01.03.2018, p. 657-669.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Multiview Boosting With Information Propagation for Classification

AU - Peng, Jing

AU - Aved, Alex J.

AU - Seetharaman, Guna

AU - Palaniappan, Kannappan

PY - 2018/3/1

Y1 - 2018/3/1

N2 - Multiview learning has shown promising potential in many applications. However, most techniques are focused on either view consistency, or view diversity. In this paper, we introduce a novel multiview boosting algorithm, called Boost.SH, that computes weak classifiers independently of each view but uses a shared weight distribution to propagate information among the multiple views to ensure consistency. To encourage diversity, we introduce randomized Boost.SH and show its convergence to the greedy Boost.SH solution in the sense of minimizing regret using the framework of adversarial multiarmed bandits. We also introduce a variant of Boost.SH that combines decisions from multiple experts for recommending views for classification. We propose an expert strategy for multiview learning based on inverse variance, which explores both consistency and diversity. Experiments on biometric recognition, document categorization, multilingual text, and yeast genomic multiview data sets demonstrate the advantage of Boost.SH (85%) compared with other boosting algorithms like AdaBoost (82%) using concatenated views and substantially better than a multiview kernel learning algorithm (74%).

AB - Multiview learning has shown promising potential in many applications. However, most techniques are focused on either view consistency, or view diversity. In this paper, we introduce a novel multiview boosting algorithm, called Boost.SH, that computes weak classifiers independently of each view but uses a shared weight distribution to propagate information among the multiple views to ensure consistency. To encourage diversity, we introduce randomized Boost.SH and show its convergence to the greedy Boost.SH solution in the sense of minimizing regret using the framework of adversarial multiarmed bandits. We also introduce a variant of Boost.SH that combines decisions from multiple experts for recommending views for classification. We propose an expert strategy for multiview learning based on inverse variance, which explores both consistency and diversity. Experiments on biometric recognition, document categorization, multilingual text, and yeast genomic multiview data sets demonstrate the advantage of Boost.SH (85%) compared with other boosting algorithms like AdaBoost (82%) using concatenated views and substantially better than a multiview kernel learning algorithm (74%).

KW - Biometrics

KW - boosting

KW - classification

KW - convergence

KW - data fusion

KW - multiarmed bandits

KW - multiview learning

UR - http://www.scopus.com/inward/record.url?scp=85008461969&partnerID=8YFLogxK

U2 - 10.1109/TNNLS.2016.2637881

DO - 10.1109/TNNLS.2016.2637881

M3 - Article

C2 - 28060713

AN - SCOPUS:85008461969

VL - 29

SP - 657

EP - 669

JO - IEEE Transactions on Neural Networks and Learning Systems

JF - IEEE Transactions on Neural Networks and Learning Systems

SN - 2162-237X

IS - 3

ER -