Towards a unified account of supervised and unsupervised category learning

Todd M. Gureckis, Bradley C. Love

Research output: Contribution to journalArticle

Abstract

(Supervised and Unsupervised STratified Adaptive IncrementalNetwork) is a network model of human category learning. SUSTAIN initially assumes a simple category structure. If simple solutions prove inadequate and SUSTAIN is confronted with a surprising event (e.g. it is told that a bat is a mammal instead of a bird), SUSTAIN recruits an additional cluster to represent the surprising event. Newly recruited clusters are available to explain future events and can themselves evolve into prototypes/attractors/rules. SUSTAIN has expanded the scope of findings that models of human category learning can address. This paper extends SUSTAIN so that it can be used to account for both supervised and unsupervised learning data through a common mechanism. A modified recruitment rule is introduced that creates new conceptual clusters in response to surprising events during learning. The new formulation of the model is called uSUSTAIN for 'unified SUSTAIN.' The implications of using a unified recruitment method for both supervised and unsupervised learning are discussed.

Original languageEnglish (US)
Pages (from-to)1-24
Number of pages24
JournalJournal of Experimental and Theoretical Artificial Intelligence
Volume15
Issue number1
DOIs
StatePublished - Jan 2003

Fingerprint

Unsupervised learning
Supervised learning
Unsupervised Learning
Supervised Learning
Mammals
Birds
Network Model
Attractor
Prototype
Learning
Formulation
Model
Human

Keywords

  • Category
  • Connectionist
  • Learning
  • Psychology
  • Supervised
  • Unsupervised

ASJC Scopus subject areas

  • Artificial Intelligence

Cite this

Towards a unified account of supervised and unsupervised category learning. / Gureckis, Todd M.; Love, Bradley C.

In: Journal of Experimental and Theoretical Artificial Intelligence, Vol. 15, No. 1, 01.2003, p. 1-24.

Research output: Contribution to journalArticle

@article{184cb2990f2948d4bc5b6246de559e64,
title = "Towards a unified account of supervised and unsupervised category learning",
abstract = "(Supervised and Unsupervised STratified Adaptive IncrementalNetwork) is a network model of human category learning. SUSTAIN initially assumes a simple category structure. If simple solutions prove inadequate and SUSTAIN is confronted with a surprising event (e.g. it is told that a bat is a mammal instead of a bird), SUSTAIN recruits an additional cluster to represent the surprising event. Newly recruited clusters are available to explain future events and can themselves evolve into prototypes/attractors/rules. SUSTAIN has expanded the scope of findings that models of human category learning can address. This paper extends SUSTAIN so that it can be used to account for both supervised and unsupervised learning data through a common mechanism. A modified recruitment rule is introduced that creates new conceptual clusters in response to surprising events during learning. The new formulation of the model is called uSUSTAIN for 'unified SUSTAIN.' The implications of using a unified recruitment method for both supervised and unsupervised learning are discussed.",
keywords = "Category, Connectionist, Learning, Psychology, Supervised, Unsupervised",
author = "Gureckis, {Todd M.} and Love, {Bradley C.}",
year = "2003",
month = "1",
doi = "10.1080/09528130210166097",
language = "English (US)",
volume = "15",
pages = "1--24",
journal = "Journal of Experimental and Theoretical Artificial Intelligence",
issn = "0952-813X",
publisher = "Taylor and Francis Ltd.",
number = "1",

}

TY - JOUR

T1 - Towards a unified account of supervised and unsupervised category learning

AU - Gureckis, Todd M.

AU - Love, Bradley C.

PY - 2003/1

Y1 - 2003/1

N2 - (Supervised and Unsupervised STratified Adaptive IncrementalNetwork) is a network model of human category learning. SUSTAIN initially assumes a simple category structure. If simple solutions prove inadequate and SUSTAIN is confronted with a surprising event (e.g. it is told that a bat is a mammal instead of a bird), SUSTAIN recruits an additional cluster to represent the surprising event. Newly recruited clusters are available to explain future events and can themselves evolve into prototypes/attractors/rules. SUSTAIN has expanded the scope of findings that models of human category learning can address. This paper extends SUSTAIN so that it can be used to account for both supervised and unsupervised learning data through a common mechanism. A modified recruitment rule is introduced that creates new conceptual clusters in response to surprising events during learning. The new formulation of the model is called uSUSTAIN for 'unified SUSTAIN.' The implications of using a unified recruitment method for both supervised and unsupervised learning are discussed.

AB - (Supervised and Unsupervised STratified Adaptive IncrementalNetwork) is a network model of human category learning. SUSTAIN initially assumes a simple category structure. If simple solutions prove inadequate and SUSTAIN is confronted with a surprising event (e.g. it is told that a bat is a mammal instead of a bird), SUSTAIN recruits an additional cluster to represent the surprising event. Newly recruited clusters are available to explain future events and can themselves evolve into prototypes/attractors/rules. SUSTAIN has expanded the scope of findings that models of human category learning can address. This paper extends SUSTAIN so that it can be used to account for both supervised and unsupervised learning data through a common mechanism. A modified recruitment rule is introduced that creates new conceptual clusters in response to surprising events during learning. The new formulation of the model is called uSUSTAIN for 'unified SUSTAIN.' The implications of using a unified recruitment method for both supervised and unsupervised learning are discussed.

KW - Category

KW - Connectionist

KW - Learning

KW - Psychology

KW - Supervised

KW - Unsupervised

UR - http://www.scopus.com/inward/record.url?scp=0038673077&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0038673077&partnerID=8YFLogxK

U2 - 10.1080/09528130210166097

DO - 10.1080/09528130210166097

M3 - Article

VL - 15

SP - 1

EP - 24

JO - Journal of Experimental and Theoretical Artificial Intelligence

JF - Journal of Experimental and Theoretical Artificial Intelligence

SN - 0952-813X

IS - 1

ER -