Context dependent learning in neural networks

L.J. Spreeuwers; B.J. van der Zwaag; F. van der Heijden

doi:10.1049/cp:19950736

Context dependent learning in neural networks

L.J. Spreeuwers, B.J. van der Zwaag, F. van der Heijden

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

1 Citation (Scopus)

158 Downloads (Pure)

Abstract

In this paper an extension to the standard error backpropagation learning rule for multi-layer feed forward neural networks is proposed, that enables them to be trained for context dependent information. The context dependent learning is realised by using a different error function (called Average Risk: AVR) in stead of the sum of squared errors (SQE) normally used in error backpropagation and by adapting the update rules. It is shown that for applications where this context dependent information is important, a major improvement in performance is obtained.

Original language	English
Title of host publication	Fifth International Conference on Image Processing and Its Applications 1995
Place of Publication	Piscataway, NJ
Publisher	IEEE
Pages	632-636
Number of pages	5
ISBN (Print)	0-85296-642-3
DOIs	https://doi.org/10.1049/cp:19950736
Publication status	Published - 1995
Event	5th International Conference on Image Processing and its Applications 1995 - Edinburgh, United Kingdom Duration: 4 Jul 1995 → 6 Jul 1995 Conference number: 5

Conference

Conference	5th International Conference on Image Processing and its Applications 1995
Country/Territory	United Kingdom
City	Edinburgh
Period	4/07/95 → 6/07/95

Keywords

SCS-Safety
AVR
Back propagation
Multilayer feedforward neural networks
Multilayer perceptrons
Neural networks
Average risk
Context-dependent learning
Feedforward neural nets
Error backpropagation
Error backpropagation learning rule
Error function

Access to Document

10.1049/cp:19950736

Spreeuwen1995contextFinal published version, 439 KB

Cite this

@inproceedings{3e7f69ba76cf48de95144816e4ea081d,

title = "Context dependent learning in neural networks",

abstract = "In this paper an extension to the standard error backpropagation learning rule for multi-layer feed forward neural networks is proposed, that enables them to be trained for context dependent information. The context dependent learning is realised by using a different error function (called Average Risk: AVR) in stead of the sum of squared errors (SQE) normally used in error backpropagation and by adapting the update rules. It is shown that for applications where this context dependent information is important, a major improvement in performance is obtained.",

keywords = "SCS-Safety, AVR, Back propagation, Multilayer feedforward neural networks, Multilayer perceptrons, Neural networks, Average risk, Context-dependent learning, Feedforward neural nets, Error backpropagation, Error backpropagation learning rule, Error function",

author = "L.J. Spreeuwers and {van der Zwaag}, B.J. and {van der Heijden}, F.",

year = "1995",

doi = "10.1049/cp:19950736",

language = "English",

isbn = "0-85296-642-3",

pages = "632--636",

booktitle = "Fifth International Conference on Image Processing and Its Applications 1995",

publisher = "IEEE",

address = "United States",

note = "5th International Conference on Image Processing and its Applications 1995 ; Conference date: 04-07-1995 Through 06-07-1995",

}

Spreeuwers, LJ , van der Zwaag, BJ & van der Heijden, F 1995, Context dependent learning in neural networks. in Fifth International Conference on Image Processing and Its Applications 1995. IEEE, Piscataway, NJ, pp. 632-636, 5th International Conference on Image Processing and its Applications 1995, Edinburgh, United Kingdom, 4/07/95. https://doi.org/10.1049/cp:19950736

TY - GEN

T1 - Context dependent learning in neural networks

AU - Spreeuwers, L.J.

AU - van der Zwaag, B.J.

AU - van der Heijden, F.

N1 - Conference code: 5

PY - 1995

Y1 - 1995

N2 - In this paper an extension to the standard error backpropagation learning rule for multi-layer feed forward neural networks is proposed, that enables them to be trained for context dependent information. The context dependent learning is realised by using a different error function (called Average Risk: AVR) in stead of the sum of squared errors (SQE) normally used in error backpropagation and by adapting the update rules. It is shown that for applications where this context dependent information is important, a major improvement in performance is obtained.

AB - In this paper an extension to the standard error backpropagation learning rule for multi-layer feed forward neural networks is proposed, that enables them to be trained for context dependent information. The context dependent learning is realised by using a different error function (called Average Risk: AVR) in stead of the sum of squared errors (SQE) normally used in error backpropagation and by adapting the update rules. It is shown that for applications where this context dependent information is important, a major improvement in performance is obtained.

KW - SCS-Safety

KW - AVR

KW - Back propagation

KW - Multilayer feedforward neural networks

KW - Multilayer perceptrons

KW - Neural networks

KW - Average risk

KW - Context-dependent learning

KW - Feedforward neural nets

KW - Error backpropagation

KW - Error backpropagation learning rule

KW - Error function

UR - http://www.scopus.com/inward/record.url?eid=2-s2.0-0029215118&partnerID=MN8TOARS

U2 - 10.1049/cp:19950736

DO - 10.1049/cp:19950736

M3 - Conference contribution

SN - 0-85296-642-3

SP - 632

EP - 636

BT - Fifth International Conference on Image Processing and Its Applications 1995

PB - IEEE

CY - Piscataway, NJ

T2 - 5th International Conference on Image Processing and its Applications 1995

Y2 - 4 July 1995 through 6 July 1995

ER -

Context dependent learning in neural networks

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this