A projection based learning meta-cognitive RBF network classifier for effective diagnosis of Parkinson's disease

G. Sateesh Babu, S. Suresh, K. Uma Sangumathi, Hyong Joong Kim

Research output: Chapter in Book/Report/Conference proceedingConference contribution

8 Citations (Scopus)

Abstract

In this paper, we proposed a 'Projection Based Learning for Meta-cognitive Radial Basis Function Network (PBL-McRBFN)' classifier for effective diagnosis of Parkinson's disease. McRBFN is inspired by human meta-cognitive learning principles. McRBFN uses the estimated class label, the maximum hinge error and class-wise significance to address the self-regulating principles of what-to-learn, when-to-learn and how-to-learn in a meta-cognitive framework. Initially, McRBFN begins with zero hidden neurons and adds required number of neurons to approximate the decision surface. When a neuron is added, network parameters are initialized based on the sample overlapping conditions. The output weights are updated using a PBL algorithm such that the network finds the minimum point of an energy function defined by the hinge-loss error. The experimental results on parkinson's data sets based on vocal and gait features clearly highlight the superior performance of PBL-McRBFN classifier over results reported in the literature for detection of individual with or without PD.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages611-620
Number of pages10
Volume7368 LNCS
EditionPART 2
DOIs
Publication statusPublished - 2012 Aug 23
Event9th International Symposium on Neural Networks, ISNN 2012 - Shenyang, China
Duration: 2012 Jul 112012 Jul 14

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 2
Volume7368 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other9th International Symposium on Neural Networks, ISNN 2012
CountryChina
CityShenyang
Period12/7/1112/7/14

Fingerprint

RBF Network
Parkinson's Disease
Radial basis function networks
Neurons
Neuron
Classifiers
Radial Basis Function Network
Classifier
Projection
Hinges
Problem-based Learning
Gait
Energy Function
Overlapping
Labels
Output
Zero
Experimental Results
Learning
Class

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Sateesh Babu, G., Suresh, S., Uma Sangumathi, K., & Kim, H. J. (2012). A projection based learning meta-cognitive RBF network classifier for effective diagnosis of Parkinson's disease. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (PART 2 ed., Vol. 7368 LNCS, pp. 611-620). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7368 LNCS, No. PART 2). https://doi.org/10.1007/978-3-642-31362-2_67

A projection based learning meta-cognitive RBF network classifier for effective diagnosis of Parkinson's disease. / Sateesh Babu, G.; Suresh, S.; Uma Sangumathi, K.; Kim, Hyong Joong.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7368 LNCS PART 2. ed. 2012. p. 611-620 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7368 LNCS, No. PART 2).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Sateesh Babu, G, Suresh, S, Uma Sangumathi, K & Kim, HJ 2012, A projection based learning meta-cognitive RBF network classifier for effective diagnosis of Parkinson's disease. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). PART 2 edn, vol. 7368 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 2, vol. 7368 LNCS, pp. 611-620, 9th International Symposium on Neural Networks, ISNN 2012, Shenyang, China, 12/7/11. https://doi.org/10.1007/978-3-642-31362-2_67
Sateesh Babu G, Suresh S, Uma Sangumathi K, Kim HJ. A projection based learning meta-cognitive RBF network classifier for effective diagnosis of Parkinson's disease. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). PART 2 ed. Vol. 7368 LNCS. 2012. p. 611-620. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2). https://doi.org/10.1007/978-3-642-31362-2_67
Sateesh Babu, G. ; Suresh, S. ; Uma Sangumathi, K. ; Kim, Hyong Joong. / A projection based learning meta-cognitive RBF network classifier for effective diagnosis of Parkinson's disease. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7368 LNCS PART 2. ed. 2012. pp. 611-620 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2).
@inproceedings{0504ef88cd7246eaa5743272c91ac67f,
title = "A projection based learning meta-cognitive RBF network classifier for effective diagnosis of Parkinson's disease",
abstract = "In this paper, we proposed a 'Projection Based Learning for Meta-cognitive Radial Basis Function Network (PBL-McRBFN)' classifier for effective diagnosis of Parkinson's disease. McRBFN is inspired by human meta-cognitive learning principles. McRBFN uses the estimated class label, the maximum hinge error and class-wise significance to address the self-regulating principles of what-to-learn, when-to-learn and how-to-learn in a meta-cognitive framework. Initially, McRBFN begins with zero hidden neurons and adds required number of neurons to approximate the decision surface. When a neuron is added, network parameters are initialized based on the sample overlapping conditions. The output weights are updated using a PBL algorithm such that the network finds the minimum point of an energy function defined by the hinge-loss error. The experimental results on parkinson's data sets based on vocal and gait features clearly highlight the superior performance of PBL-McRBFN classifier over results reported in the literature for detection of individual with or without PD.",
author = "{Sateesh Babu}, G. and S. Suresh and {Uma Sangumathi}, K. and Kim, {Hyong Joong}",
year = "2012",
month = "8",
day = "23",
doi = "10.1007/978-3-642-31362-2_67",
language = "English",
isbn = "9783642313615",
volume = "7368 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
number = "PART 2",
pages = "611--620",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
edition = "PART 2",

}

TY - GEN

T1 - A projection based learning meta-cognitive RBF network classifier for effective diagnosis of Parkinson's disease

AU - Sateesh Babu, G.

AU - Suresh, S.

AU - Uma Sangumathi, K.

AU - Kim, Hyong Joong

PY - 2012/8/23

Y1 - 2012/8/23

N2 - In this paper, we proposed a 'Projection Based Learning for Meta-cognitive Radial Basis Function Network (PBL-McRBFN)' classifier for effective diagnosis of Parkinson's disease. McRBFN is inspired by human meta-cognitive learning principles. McRBFN uses the estimated class label, the maximum hinge error and class-wise significance to address the self-regulating principles of what-to-learn, when-to-learn and how-to-learn in a meta-cognitive framework. Initially, McRBFN begins with zero hidden neurons and adds required number of neurons to approximate the decision surface. When a neuron is added, network parameters are initialized based on the sample overlapping conditions. The output weights are updated using a PBL algorithm such that the network finds the minimum point of an energy function defined by the hinge-loss error. The experimental results on parkinson's data sets based on vocal and gait features clearly highlight the superior performance of PBL-McRBFN classifier over results reported in the literature for detection of individual with or without PD.

AB - In this paper, we proposed a 'Projection Based Learning for Meta-cognitive Radial Basis Function Network (PBL-McRBFN)' classifier for effective diagnosis of Parkinson's disease. McRBFN is inspired by human meta-cognitive learning principles. McRBFN uses the estimated class label, the maximum hinge error and class-wise significance to address the self-regulating principles of what-to-learn, when-to-learn and how-to-learn in a meta-cognitive framework. Initially, McRBFN begins with zero hidden neurons and adds required number of neurons to approximate the decision surface. When a neuron is added, network parameters are initialized based on the sample overlapping conditions. The output weights are updated using a PBL algorithm such that the network finds the minimum point of an energy function defined by the hinge-loss error. The experimental results on parkinson's data sets based on vocal and gait features clearly highlight the superior performance of PBL-McRBFN classifier over results reported in the literature for detection of individual with or without PD.

UR - http://www.scopus.com/inward/record.url?scp=84865109404&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84865109404&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-31362-2_67

DO - 10.1007/978-3-642-31362-2_67

M3 - Conference contribution

AN - SCOPUS:84865109404

SN - 9783642313615

VL - 7368 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 611

EP - 620

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

ER -