Φ-divergence loss-based artificial neural network

R. L. Salamwade; D. M. Sakate; S. K. Mathur

doi:10.22237/jmasm/1551966252

Φ-divergence loss-based artificial neural network

R. L. Salamwade, D. M. Sakate, S. K. Mathur

Research output: Contribution to journal › Article › peer-review

Abstract

Artificial Neural Networks (ANNs) can fit non-linear functions and recognize patterns better than several standard techniques. Performance of ANNs is measured by using loss functions. Phi-divergence estimator is generalization of maximum likelihood estimator and it possesses all its properties. A neural network is proposed which is trained using phidivergence loss.

Original language	English (US)
Article number	eP2646
Journal	Journal of Modern Applied Statistical Methods
Volume	17
Issue number	2
DOIs	https://doi.org/10.22237/jmasm/1551966252
State	Published - 2018

Keywords

Back-propagation
Classification
Loss function
Mean square error
Power divergence family

ASJC Scopus subject areas

Statistics and Probability
Statistics, Probability and Uncertainty

Access to Document

10.22237/jmasm/1551966252

Cite this

@article{c2a73f99040f4c1ab703a6a47e3d51e5,

title = "Φ-divergence loss-based artificial neural network",

abstract = "Artificial Neural Networks (ANNs) can fit non-linear functions and recognize patterns better than several standard techniques. Performance of ANNs is measured by using loss functions. Phi-divergence estimator is generalization of maximum likelihood estimator and it possesses all its properties. A neural network is proposed which is trained using phidivergence loss.",

keywords = "Back-propagation, Classification, Loss function, Mean square error, Power divergence family",

author = "Salamwade, {R. L.} and Sakate, {D. M.} and Mathur, {S. K.}",

note = "Funding Information: The first and second author wish to thank the Department of Science and Technology, New Delhi, India for financial support under DST-PURSE scheme. Publisher Copyright: {\textcopyright} 2018 Wayne State University.",

year = "2018",

doi = "10.22237/jmasm/1551966252",

language = "English (US)",

volume = "17",

journal = "Journal of Modern Applied Statistical Methods",

issn = "1538-9472",

publisher = "Wayne State University",

number = "2",

}

TY - JOUR

T1 - Φ-divergence loss-based artificial neural network

AU - Salamwade, R. L.

AU - Sakate, D. M.

AU - Mathur, S. K.

N1 - Funding Information: The first and second author wish to thank the Department of Science and Technology, New Delhi, India for financial support under DST-PURSE scheme. Publisher Copyright: © 2018 Wayne State University.

PY - 2018

Y1 - 2018

N2 - Artificial Neural Networks (ANNs) can fit non-linear functions and recognize patterns better than several standard techniques. Performance of ANNs is measured by using loss functions. Phi-divergence estimator is generalization of maximum likelihood estimator and it possesses all its properties. A neural network is proposed which is trained using phidivergence loss.

AB - Artificial Neural Networks (ANNs) can fit non-linear functions and recognize patterns better than several standard techniques. Performance of ANNs is measured by using loss functions. Phi-divergence estimator is generalization of maximum likelihood estimator and it possesses all its properties. A neural network is proposed which is trained using phidivergence loss.

KW - Back-propagation

KW - Classification

KW - Loss function

KW - Mean square error

KW - Power divergence family

UR - http://www.scopus.com/inward/record.url?scp=85063735337&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85063735337&partnerID=8YFLogxK

U2 - 10.22237/jmasm/1551966252

DO - 10.22237/jmasm/1551966252

M3 - Article

AN - SCOPUS:85063735337

SN - 1538-9472

VL - 17

JO - Journal of Modern Applied Statistical Methods

JF - Journal of Modern Applied Statistical Methods

IS - 2

M1 - eP2646

ER -

Φ-divergence loss-based artificial neural network

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this