On the feature selection criterion based on an approximation of multidimensional mutual information

Kiran S. Balagani; Vir V. Phoha

doi:10.1109/TPAMI.2010.62

On the feature selection criterion based on an approximation of multidimensional mutual information

Kiran S. Balagani, Vir V. Phoha

Research output: Contribution to journal › Review article › peer-review

44 Scopus citations

Abstract

We derive the feature selection criterion presented in [1] and [2] from the multidimensional mutual information between features and the class. Our derivation: 1) specifies and validates the lower-order dependency assumptions of the criterion and 2) mathematically justifies the utility of the criterion by relating it to Bayes classification error.

Original language	English (US)
Article number	5432207
Pages (from-to)	1342-1343
Number of pages	2
Journal	IEEE Transactions on Pattern Analysis and Machine Intelligence
Volume	32
Issue number	7
DOIs	https://doi.org/10.1109/TPAMI.2010.62
State	Published - 2010
Externally published	Yes

Keywords

Bayes classification error
Entropy
Entropy estimation
Feature selection
Mutual information

ASJC Scopus subject areas

Software
Computer Vision and Pattern Recognition
Computational Theory and Mathematics
Artificial Intelligence
Applied Mathematics

Access to Document

10.1109/TPAMI.2010.62

Cite this

@article{328f9abed561442dbdc1e73563499bee,

title = "On the feature selection criterion based on an approximation of multidimensional mutual information",

abstract = "We derive the feature selection criterion presented in [1] and [2] from the multidimensional mutual information between features and the class. Our derivation: 1) specifies and validates the lower-order dependency assumptions of the criterion and 2) mathematically justifies the utility of the criterion by relating it to Bayes classification error.",

keywords = "Bayes classification error, Entropy, Entropy estimation, Feature selection, Mutual information",

author = "Balagani, {Kiran S.} and Phoha, {Vir V.}",

note = "Funding Information: The authors thank the three anonymous reviewers for their comments. In particular, they are thankful to Reviewer 1 and Reviewer 2 for their insightful suggestions that improved the paper. This work was supported in part by the Louisiana Board of Regents under P-KSFI Grant LEQSF (2007-12)-ENH-PKSFI-PRS-03.",

year = "2010",

doi = "10.1109/TPAMI.2010.62",

language = "English (US)",

volume = "32",

pages = "1342--1343",

journal = "IEEE Transactions on Pattern Analysis and Machine Intelligence",

issn = "0162-8828",

publisher = "IEEE Computer Society",

number = "7",

}

TY - JOUR

T1 - On the feature selection criterion based on an approximation of multidimensional mutual information

AU - Balagani, Kiran S.

AU - Phoha, Vir V.

N1 - Funding Information: The authors thank the three anonymous reviewers for their comments. In particular, they are thankful to Reviewer 1 and Reviewer 2 for their insightful suggestions that improved the paper. This work was supported in part by the Louisiana Board of Regents under P-KSFI Grant LEQSF (2007-12)-ENH-PKSFI-PRS-03.

PY - 2010

Y1 - 2010

N2 - We derive the feature selection criterion presented in [1] and [2] from the multidimensional mutual information between features and the class. Our derivation: 1) specifies and validates the lower-order dependency assumptions of the criterion and 2) mathematically justifies the utility of the criterion by relating it to Bayes classification error.

AB - We derive the feature selection criterion presented in [1] and [2] from the multidimensional mutual information between features and the class. Our derivation: 1) specifies and validates the lower-order dependency assumptions of the criterion and 2) mathematically justifies the utility of the criterion by relating it to Bayes classification error.

KW - Bayes classification error

KW - Entropy

KW - Entropy estimation

KW - Feature selection

KW - Mutual information

UR - http://www.scopus.com/inward/record.url?scp=77952590681&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77952590681&partnerID=8YFLogxK

U2 - 10.1109/TPAMI.2010.62

DO - 10.1109/TPAMI.2010.62

M3 - Review article

C2 - 20489237

AN - SCOPUS:77952590681

SN - 0162-8828

VL - 32

SP - 1342

EP - 1343

JO - IEEE Transactions on Pattern Analysis and Machine Intelligence

JF - IEEE Transactions on Pattern Analysis and Machine Intelligence

IS - 7

M1 - 5432207

ER -

On the feature selection criterion based on an approximation of multidimensional mutual information

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this