Comparing generalization by humans and adaptive networks

M. Pavel, Mark A. Gluck, Van Henkle

Research output: Contribution to journalArticle

Abstract

One goal of study reported was to examine how people generalize in a simple deterministic categorization task in which each pattern is characterized in terms of known binary features. While we expected certain similarities to emerge across human learners we anticipated that the particular generalizations might be subject to considerable individual differences. To test this idea, we used an experimental paradigm that would permit us to to observe individual subjects during the learning of a categorization task on a set of training patterns and then allow us examine the types of categorizations they made on a set of novel test patterns. Later we compared human generalizations to those of a small adaptive network. We have demonstrated that subjects who learn the same pattern categorization may abstract different principles and therefore show large individual differences in their generalization behavior. Adaptive networks with the minimum number of hidden units exhibit a similar behavior but generalize differently.

Original languageEnglish (US)
Pages (from-to)208
Number of pages1
JournalNeural Networks
Volume1
Issue number1 SUPPL
DOIs
StatePublished - 1988
Externally publishedYes

Fingerprint

Individuality
Learning

ASJC Scopus subject areas

  • Artificial Intelligence
  • Neuroscience(all)

Cite this

Comparing generalization by humans and adaptive networks. / Pavel, M.; Gluck, Mark A.; Henkle, Van.

In: Neural Networks, Vol. 1, No. 1 SUPPL, 1988, p. 208.

Research output: Contribution to journalArticle

Pavel, M. ; Gluck, Mark A. ; Henkle, Van. / Comparing generalization by humans and adaptive networks. In: Neural Networks. 1988 ; Vol. 1, No. 1 SUPPL. pp. 208.
@article{b49ac79dcd5f45abb1d9464038e58f8c,
title = "Comparing generalization by humans and adaptive networks",
abstract = "One goal of study reported was to examine how people generalize in a simple deterministic categorization task in which each pattern is characterized in terms of known binary features. While we expected certain similarities to emerge across human learners we anticipated that the particular generalizations might be subject to considerable individual differences. To test this idea, we used an experimental paradigm that would permit us to to observe individual subjects during the learning of a categorization task on a set of training patterns and then allow us examine the types of categorizations they made on a set of novel test patterns. Later we compared human generalizations to those of a small adaptive network. We have demonstrated that subjects who learn the same pattern categorization may abstract different principles and therefore show large individual differences in their generalization behavior. Adaptive networks with the minimum number of hidden units exhibit a similar behavior but generalize differently.",
author = "M. Pavel and Gluck, {Mark A.} and Van Henkle",
year = "1988",
doi = "10.1016/0893-6080(88)90245-6",
language = "English (US)",
volume = "1",
pages = "208",
journal = "Neural Networks",
issn = "0893-6080",
publisher = "Elsevier Limited",
number = "1 SUPPL",

}

TY - JOUR

T1 - Comparing generalization by humans and adaptive networks

AU - Pavel, M.

AU - Gluck, Mark A.

AU - Henkle, Van

PY - 1988

Y1 - 1988

N2 - One goal of study reported was to examine how people generalize in a simple deterministic categorization task in which each pattern is characterized in terms of known binary features. While we expected certain similarities to emerge across human learners we anticipated that the particular generalizations might be subject to considerable individual differences. To test this idea, we used an experimental paradigm that would permit us to to observe individual subjects during the learning of a categorization task on a set of training patterns and then allow us examine the types of categorizations they made on a set of novel test patterns. Later we compared human generalizations to those of a small adaptive network. We have demonstrated that subjects who learn the same pattern categorization may abstract different principles and therefore show large individual differences in their generalization behavior. Adaptive networks with the minimum number of hidden units exhibit a similar behavior but generalize differently.

AB - One goal of study reported was to examine how people generalize in a simple deterministic categorization task in which each pattern is characterized in terms of known binary features. While we expected certain similarities to emerge across human learners we anticipated that the particular generalizations might be subject to considerable individual differences. To test this idea, we used an experimental paradigm that would permit us to to observe individual subjects during the learning of a categorization task on a set of training patterns and then allow us examine the types of categorizations they made on a set of novel test patterns. Later we compared human generalizations to those of a small adaptive network. We have demonstrated that subjects who learn the same pattern categorization may abstract different principles and therefore show large individual differences in their generalization behavior. Adaptive networks with the minimum number of hidden units exhibit a similar behavior but generalize differently.

UR - http://www.scopus.com/inward/record.url?scp=0024167312&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0024167312&partnerID=8YFLogxK

U2 - 10.1016/0893-6080(88)90245-6

DO - 10.1016/0893-6080(88)90245-6

M3 - Article

AN - SCOPUS:0024167312

VL - 1

SP - 208

JO - Neural Networks

JF - Neural Networks

SN - 0893-6080

IS - 1 SUPPL

ER -