Discriminative concept learning network: Reveal high-level differential concepts from shallow architecture

Q. Wang, Sylvia Young, A. Harwood, C.S. Ong

Research output: Chapter in Book/Conference paperConference paper

1 Citation (Scopus)

Abstract

© 2015 IEEE. A desired capability of deep learning is to understand the high-level, class-specific features via hierarchical features learning. However the training of deep architectures is costly comparing to simple shallow models. Bringing the high-level feature understanding into a simple shallow architecture remains an open question. We proposed a supervised learning algorithm, enabling binary classification along with an intrinsic ability of learning high-level discriminative concepts via a shallow neural network architecture. The physical architecture of the network has one hidden layer (also serving as the output layer) responsible for the classification and an input layer directly identifies the informative features that constitute the high-level differential concepts between the two classes. Compared to other shallow classifiers, we demonstrate its practicability in real world classification problems. We also illustrate the human-understandable, discriminative concepts learned from the two image recognition exercises. Lastly, we show how it is useful in validating the disease-associated genetic variants in human genome as a real diagnostic genomics application.
Original languageEnglish
Title of host publication2015 International Joint Conference on Neural Networks (IJCNN)
Place of PublicationIreland
PublisherIEEE, Institute of Electrical and Electronics Engineers
Pages1-9
Volume2015
ISBN (Print)9781479919604
DOIs
Publication statusPublished - 2015
Event2015 International Joint Conference on Neural Networks - Killarney, Ireland
Duration: 12 Jul 201517 Jul 2015

Conference

Conference2015 International Joint Conference on Neural Networks
CountryIreland
CityKillarney
Period12/07/1517/07/15

Fingerprint

Image recognition
Supervised learning
Network architecture
Learning algorithms
Classifiers
Genes
Neural networks
Deep learning
Genomics

Cite this

Wang, Q., Young, S., Harwood, A., & Ong, C. S. (2015). Discriminative concept learning network: Reveal high-level differential concepts from shallow architecture. In 2015 International Joint Conference on Neural Networks (IJCNN) (Vol. 2015, pp. 1-9). Ireland: IEEE, Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/IJCNN.2015.7280525
Wang, Q. ; Young, Sylvia ; Harwood, A. ; Ong, C.S. / Discriminative concept learning network: Reveal high-level differential concepts from shallow architecture. 2015 International Joint Conference on Neural Networks (IJCNN). Vol. 2015 Ireland : IEEE, Institute of Electrical and Electronics Engineers, 2015. pp. 1-9
@inproceedings{ddb7db51ca2f4333ab6220fa06f65689,
title = "Discriminative concept learning network: Reveal high-level differential concepts from shallow architecture",
abstract = "{\circledC} 2015 IEEE. A desired capability of deep learning is to understand the high-level, class-specific features via hierarchical features learning. However the training of deep architectures is costly comparing to simple shallow models. Bringing the high-level feature understanding into a simple shallow architecture remains an open question. We proposed a supervised learning algorithm, enabling binary classification along with an intrinsic ability of learning high-level discriminative concepts via a shallow neural network architecture. The physical architecture of the network has one hidden layer (also serving as the output layer) responsible for the classification and an input layer directly identifies the informative features that constitute the high-level differential concepts between the two classes. Compared to other shallow classifiers, we demonstrate its practicability in real world classification problems. We also illustrate the human-understandable, discriminative concepts learned from the two image recognition exercises. Lastly, we show how it is useful in validating the disease-associated genetic variants in human genome as a real diagnostic genomics application.",
author = "Q. Wang and Sylvia Young and A. Harwood and C.S. Ong",
year = "2015",
doi = "10.1109/IJCNN.2015.7280525",
language = "English",
isbn = "9781479919604",
volume = "2015",
pages = "1--9",
booktitle = "2015 International Joint Conference on Neural Networks (IJCNN)",
publisher = "IEEE, Institute of Electrical and Electronics Engineers",
address = "United States",

}

Wang, Q, Young, S, Harwood, A & Ong, CS 2015, Discriminative concept learning network: Reveal high-level differential concepts from shallow architecture. in 2015 International Joint Conference on Neural Networks (IJCNN). vol. 2015, IEEE, Institute of Electrical and Electronics Engineers, Ireland, pp. 1-9, 2015 International Joint Conference on Neural Networks, Killarney, Ireland, 12/07/15. https://doi.org/10.1109/IJCNN.2015.7280525

Discriminative concept learning network: Reveal high-level differential concepts from shallow architecture. / Wang, Q.; Young, Sylvia; Harwood, A.; Ong, C.S.

2015 International Joint Conference on Neural Networks (IJCNN). Vol. 2015 Ireland : IEEE, Institute of Electrical and Electronics Engineers, 2015. p. 1-9.

Research output: Chapter in Book/Conference paperConference paper

TY - GEN

T1 - Discriminative concept learning network: Reveal high-level differential concepts from shallow architecture

AU - Wang, Q.

AU - Young, Sylvia

AU - Harwood, A.

AU - Ong, C.S.

PY - 2015

Y1 - 2015

N2 - © 2015 IEEE. A desired capability of deep learning is to understand the high-level, class-specific features via hierarchical features learning. However the training of deep architectures is costly comparing to simple shallow models. Bringing the high-level feature understanding into a simple shallow architecture remains an open question. We proposed a supervised learning algorithm, enabling binary classification along with an intrinsic ability of learning high-level discriminative concepts via a shallow neural network architecture. The physical architecture of the network has one hidden layer (also serving as the output layer) responsible for the classification and an input layer directly identifies the informative features that constitute the high-level differential concepts between the two classes. Compared to other shallow classifiers, we demonstrate its practicability in real world classification problems. We also illustrate the human-understandable, discriminative concepts learned from the two image recognition exercises. Lastly, we show how it is useful in validating the disease-associated genetic variants in human genome as a real diagnostic genomics application.

AB - © 2015 IEEE. A desired capability of deep learning is to understand the high-level, class-specific features via hierarchical features learning. However the training of deep architectures is costly comparing to simple shallow models. Bringing the high-level feature understanding into a simple shallow architecture remains an open question. We proposed a supervised learning algorithm, enabling binary classification along with an intrinsic ability of learning high-level discriminative concepts via a shallow neural network architecture. The physical architecture of the network has one hidden layer (also serving as the output layer) responsible for the classification and an input layer directly identifies the informative features that constitute the high-level differential concepts between the two classes. Compared to other shallow classifiers, we demonstrate its practicability in real world classification problems. We also illustrate the human-understandable, discriminative concepts learned from the two image recognition exercises. Lastly, we show how it is useful in validating the disease-associated genetic variants in human genome as a real diagnostic genomics application.

U2 - 10.1109/IJCNN.2015.7280525

DO - 10.1109/IJCNN.2015.7280525

M3 - Conference paper

SN - 9781479919604

VL - 2015

SP - 1

EP - 9

BT - 2015 International Joint Conference on Neural Networks (IJCNN)

PB - IEEE, Institute of Electrical and Electronics Engineers

CY - Ireland

ER -

Wang Q, Young S, Harwood A, Ong CS. Discriminative concept learning network: Reveal high-level differential concepts from shallow architecture. In 2015 International Joint Conference on Neural Networks (IJCNN). Vol. 2015. Ireland: IEEE, Institute of Electrical and Electronics Engineers. 2015. p. 1-9 https://doi.org/10.1109/IJCNN.2015.7280525