Learning from Multiple Imperfect Instructors in Sensor Networks

Nurali Virani, Shashi Phoha, Asok Ray

Research output: Contribution to journalArticle

Abstract

This paper presents a sequential learning framework for sensors in a network, where a few sensors assume the role of an instructor to train other sensors in the network. The instructors provide estimated labels for measurements of new sensors. These labels are possibly noisy, because a classifier of the instructor may not be perfect. A recursive density estimator is proposed to obtain the true measurement model (i.e., the observation density conditioned on the label) in spite of the training with noisy labels. Specifically, this paper answers the question "Can a sensor train other sensors?", provides necessary conditions for sensors to act as instructors, presents a sequential learning framework using recursive nonparametric kernel density estimation, and provides a convergence rate for the expected error in an observation density. The underlying concepts are illustrated and validated with simulation results.

Original languageEnglish (US)
Article number8278847
Pages (from-to)5166-5172
Number of pages7
JournalIEEE Transactions on Neural Networks and Learning Systems
Volume29
Issue number10
DOIs
StatePublished - Oct 1 2018

Fingerprint

Sensor networks
Sensors
Labels
Classifiers

All Science Journal Classification (ASJC) codes

  • Software
  • Computer Science Applications
  • Computer Networks and Communications
  • Artificial Intelligence

Cite this

@article{89f6d0820aef4ae49105aa155431164e,
title = "Learning from Multiple Imperfect Instructors in Sensor Networks",
abstract = "This paper presents a sequential learning framework for sensors in a network, where a few sensors assume the role of an instructor to train other sensors in the network. The instructors provide estimated labels for measurements of new sensors. These labels are possibly noisy, because a classifier of the instructor may not be perfect. A recursive density estimator is proposed to obtain the true measurement model (i.e., the observation density conditioned on the label) in spite of the training with noisy labels. Specifically, this paper answers the question {"}Can a sensor train other sensors?{"}, provides necessary conditions for sensors to act as instructors, presents a sequential learning framework using recursive nonparametric kernel density estimation, and provides a convergence rate for the expected error in an observation density. The underlying concepts are illustrated and validated with simulation results.",
author = "Nurali Virani and Shashi Phoha and Asok Ray",
year = "2018",
month = "10",
day = "1",
doi = "10.1109/TNNLS.2018.2791898",
language = "English (US)",
volume = "29",
pages = "5166--5172",
journal = "IEEE Transactions on Neural Networks and Learning Systems",
issn = "2162-237X",
publisher = "IEEE Computational Intelligence Society",
number = "10",

}

Learning from Multiple Imperfect Instructors in Sensor Networks. / Virani, Nurali; Phoha, Shashi; Ray, Asok.

In: IEEE Transactions on Neural Networks and Learning Systems, Vol. 29, No. 10, 8278847, 01.10.2018, p. 5166-5172.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Learning from Multiple Imperfect Instructors in Sensor Networks

AU - Virani, Nurali

AU - Phoha, Shashi

AU - Ray, Asok

PY - 2018/10/1

Y1 - 2018/10/1

N2 - This paper presents a sequential learning framework for sensors in a network, where a few sensors assume the role of an instructor to train other sensors in the network. The instructors provide estimated labels for measurements of new sensors. These labels are possibly noisy, because a classifier of the instructor may not be perfect. A recursive density estimator is proposed to obtain the true measurement model (i.e., the observation density conditioned on the label) in spite of the training with noisy labels. Specifically, this paper answers the question "Can a sensor train other sensors?", provides necessary conditions for sensors to act as instructors, presents a sequential learning framework using recursive nonparametric kernel density estimation, and provides a convergence rate for the expected error in an observation density. The underlying concepts are illustrated and validated with simulation results.

AB - This paper presents a sequential learning framework for sensors in a network, where a few sensors assume the role of an instructor to train other sensors in the network. The instructors provide estimated labels for measurements of new sensors. These labels are possibly noisy, because a classifier of the instructor may not be perfect. A recursive density estimator is proposed to obtain the true measurement model (i.e., the observation density conditioned on the label) in spite of the training with noisy labels. Specifically, this paper answers the question "Can a sensor train other sensors?", provides necessary conditions for sensors to act as instructors, presents a sequential learning framework using recursive nonparametric kernel density estimation, and provides a convergence rate for the expected error in an observation density. The underlying concepts are illustrated and validated with simulation results.

UR - http://www.scopus.com/inward/record.url?scp=85041678360&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85041678360&partnerID=8YFLogxK

U2 - 10.1109/TNNLS.2018.2791898

DO - 10.1109/TNNLS.2018.2791898

M3 - Article

C2 - 29994429

AN - SCOPUS:85041678360

VL - 29

SP - 5166

EP - 5172

JO - IEEE Transactions on Neural Networks and Learning Systems

JF - IEEE Transactions on Neural Networks and Learning Systems

SN - 2162-237X

IS - 10

M1 - 8278847

ER -