State-trace analysis of sequence learning by simple recurrent networks

Fayme Yeates; Andy Wills; Fergal Jones; Ian McLaren

State-trace analysis of sequence learning by simple recurrent networks

Fayme Yeates
, Andy Wills
, Fergal Jones
, Ian McLaren

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

1 Citation (Scopus)

Abstract

This study investigated the use of state-trace analysis (Bamber, 1979) when applied to computational models of human learning. We aimed to investigate the performance of simple recurrent networks (SRNs) on a sequence learning task. Elman’s (1990) SRN and Cleeremans & McClelland’s (1991) Augmented SRN are both benchmark models of human sequence learning. The differences between these models, comprising of an additional learning parameter and the use of response units activated by output units constituted our main manipulation. The results are presented as a state-trace analysis, which demonstrates that the addition of an additional type of weight component, and response units to a SRN produces multi-dimensional state-trace plots.
However, varying the learning rate parameter of the SRN also produced two functions on a state-trace plot, suggesting that state-trace analysis may be sensitive to variation within a single process.

Original language	English
Title of host publication	Proceedings of the 34th Annual Meeting of the Cognitive Science Society
Editors	Naomi Miyake, David Peebles, Richard P. Cooper
Place of Publication	Austin, Texas
Publisher	Cognitive Science Society
Pages	2581-2586
Number of pages	6
ISBN (Electronic)	9780976831884
ISBN (Print)	9780976831884
Publication status	Published - Aug 2012
Externally published	Yes
Event	34th Annual Meeting of the Cognitive Science Society: Building Bridges Across Cognitive Sciences Around the World, CogSci 2012 - Sapporo, Japan Duration: 1 Aug 2012 → 4 Aug 2012

Academic conference

Academic conference	34th Annual Meeting of the Cognitive Science Society: Building Bridges Across Cognitive Sciences Around the World, CogSci 2012
Country/Territory	Japan
City	Sapporo
Period	1/08/12 → 4/08/12

Keywords

Augmented SRN
Learning
sequence learning
SRN
state-trace analysis

Access to Document

https://cognitivesciencesociety.org/wp-content/uploads/2019/01/cogsci12_proceedings.pdfLicence: Unspecified

Cite this

Yeates, F., Wills, A., Jones, F., & McLaren, I. (2012). State-trace analysis of sequence learning by simple recurrent networks. In N. Miyake, D. Peebles, & R. P. Cooper (Eds.), Proceedings of the 34th Annual Meeting of the Cognitive Science Society (pp. 2581-2586). Cognitive Science Society. https://cognitivesciencesociety.org/wp-content/uploads/2019/01/cogsci12_proceedings.pdf

@inproceedings{ae0219fe433f4980b5bec7f20d9e5f42,

title = "State-trace analysis of sequence learning by simple recurrent networks",

abstract = "This study investigated the use of state-trace analysis (Bamber, 1979) when applied to computational models of human learning. We aimed to investigate the performance of simple recurrent networks (SRNs) on a sequence learning task. Elman{\textquoteright}s (1990) SRN and Cleeremans \& McClelland{\textquoteright}s (1991) Augmented SRN are both benchmark models of human sequence learning. The differences between these models, comprising of an additional learning parameter and the use of response units activated by output units constituted our main manipulation. The results are presented as a state-trace analysis, which demonstrates that the addition of an additional type of weight component, and response units to a SRN produces multi-dimensional state-trace plots. However, varying the learning rate parameter of the SRN also produced two functions on a state-trace plot, suggesting that state-trace analysis may be sensitive to variation within a single process.",

keywords = "Augmented SRN, Learning, sequence learning, SRN, state-trace analysis",

author = "Fayme Yeates and Andy Wills and Fergal Jones and Ian McLaren",

note = "Publisher Copyright: {\textcopyright} CogSci 2012.All rights reserved.; 34th Annual Meeting of the Cognitive Science Society: Building Bridges Across Cognitive Sciences Around the World, CogSci 2012 ; Conference date: 01-08-2012 Through 04-08-2012",

year = "2012",

month = aug,

language = "English",

isbn = "9780976831884",

pages = "2581--2586",

editor = "Naomi Miyake and David Peebles and Cooper, \{Richard P.\}",

booktitle = "Proceedings of the 34th Annual Meeting of the Cognitive Science Society",

publisher = "Cognitive Science Society",

address = "United States",

}

Yeates, F, Wills, A, Jones, F & McLaren, I 2012, State-trace analysis of sequence learning by simple recurrent networks. in N Miyake, D Peebles & RP Cooper (eds), Proceedings of the 34th Annual Meeting of the Cognitive Science Society. Cognitive Science Society, Austin, Texas, pp. 2581-2586, 34th Annual Meeting of the Cognitive Science Society: Building Bridges Across Cognitive Sciences Around the World, CogSci 2012, Sapporo, Japan, 1/08/12. <https://cognitivesciencesociety.org/wp-content/uploads/2019/01/cogsci12_proceedings.pdf>

State-trace analysis of sequence learning by simple recurrent networks. / Yeates, Fayme; Wills, Andy; Jones, Fergal et al.
Proceedings of the 34th Annual Meeting of the Cognitive Science Society. ed. / Naomi Miyake; David Peebles; Richard P. Cooper. Austin, Texas: Cognitive Science Society, 2012. p. 2581-2586.

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

TY - GEN

T1 - State-trace analysis of sequence learning by simple recurrent networks

AU - Yeates, Fayme

AU - Wills, Andy

AU - Jones, Fergal

AU - McLaren, Ian

PY - 2012/8

Y1 - 2012/8

N2 - This study investigated the use of state-trace analysis (Bamber, 1979) when applied to computational models of human learning. We aimed to investigate the performance of simple recurrent networks (SRNs) on a sequence learning task. Elman’s (1990) SRN and Cleeremans & McClelland’s (1991) Augmented SRN are both benchmark models of human sequence learning. The differences between these models, comprising of an additional learning parameter and the use of response units activated by output units constituted our main manipulation. The results are presented as a state-trace analysis, which demonstrates that the addition of an additional type of weight component, and response units to a SRN produces multi-dimensional state-trace plots. However, varying the learning rate parameter of the SRN also produced two functions on a state-trace plot, suggesting that state-trace analysis may be sensitive to variation within a single process.

AB - This study investigated the use of state-trace analysis (Bamber, 1979) when applied to computational models of human learning. We aimed to investigate the performance of simple recurrent networks (SRNs) on a sequence learning task. Elman’s (1990) SRN and Cleeremans & McClelland’s (1991) Augmented SRN are both benchmark models of human sequence learning. The differences between these models, comprising of an additional learning parameter and the use of response units activated by output units constituted our main manipulation. The results are presented as a state-trace analysis, which demonstrates that the addition of an additional type of weight component, and response units to a SRN produces multi-dimensional state-trace plots. However, varying the learning rate parameter of the SRN also produced two functions on a state-trace plot, suggesting that state-trace analysis may be sensitive to variation within a single process.

KW - Augmented SRN

KW - Learning

KW - sequence learning

KW - SRN

KW - state-trace analysis

UR - https://www.scopus.com/pages/publications/84936960170

M3 - Conference contribution

AN - SCOPUS:84936960170

SN - 9780976831884

SP - 2581

EP - 2586

BT - Proceedings of the 34th Annual Meeting of the Cognitive Science Society

A2 - Miyake, Naomi

A2 - Peebles, David

A2 - Cooper, Richard P.

PB - Cognitive Science Society

CY - Austin, Texas

T2 - 34th Annual Meeting of the Cognitive Science Society: Building Bridges Across Cognitive Sciences Around the World, CogSci 2012

Y2 - 1 August 2012 through 4 August 2012

ER -

State-trace analysis of sequence learning by simple recurrent networks

Abstract

Academic conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this