Category dimensionality and feature knowledge

When more features are learned as easily as fewer

Aaron B. Hoffman, Gregory L. Murphy

Research output: Contribution to journalArticle

Abstract

Three experiments compared the learning of lower-dimensional family resemblance categories (4 dimensions) with the learning of higher-dimensional ones (8 dimensions). Category-learning models incorporating error-driven learning, hypothesis testing, or limited capacity attention predict that additional dimensions should either increase learning difficulty or decrease learning of individual features. Contrary to these predictions, the experiments showed no slower learning of high-dimensional categories; instead, subjects learned more features from high-dimensional categories than from low-dimensional categories. This result obtained both in standard learning with feedback and in noncontingent, observational learning. These results show that rather than interfering with learning, categories with more dimensions cause individuals to learn more. The authors contrast the learning of family resemblance categories with learning in classical conditioning and probability learning paradigms, in which competition among features is well documented.

Original languageEnglish (US)
Pages (from-to)301-315
Number of pages15
JournalJournal of Experimental Psychology: Learning Memory and Cognition
Volume32
Issue number2
DOIs
StatePublished - Mar 2006

Fingerprint

Learning
learning
Probability Learning
Classical Conditioning
hypothesis testing
learning disorder
experiment
conditioning
paradigm
cause

Keywords

  • Category learning
  • Concepts
  • Error-driven learning
  • Unsupervised learning

ASJC Scopus subject areas

  • Psychology(all)

Cite this

Category dimensionality and feature knowledge : When more features are learned as easily as fewer. / Hoffman, Aaron B.; Murphy, Gregory L.

In: Journal of Experimental Psychology: Learning Memory and Cognition, Vol. 32, No. 2, 03.2006, p. 301-315.

Research output: Contribution to journalArticle

@article{311f1603aeb94b818be21b9c9dd82f96,
title = "Category dimensionality and feature knowledge: When more features are learned as easily as fewer",
abstract = "Three experiments compared the learning of lower-dimensional family resemblance categories (4 dimensions) with the learning of higher-dimensional ones (8 dimensions). Category-learning models incorporating error-driven learning, hypothesis testing, or limited capacity attention predict that additional dimensions should either increase learning difficulty or decrease learning of individual features. Contrary to these predictions, the experiments showed no slower learning of high-dimensional categories; instead, subjects learned more features from high-dimensional categories than from low-dimensional categories. This result obtained both in standard learning with feedback and in noncontingent, observational learning. These results show that rather than interfering with learning, categories with more dimensions cause individuals to learn more. The authors contrast the learning of family resemblance categories with learning in classical conditioning and probability learning paradigms, in which competition among features is well documented.",
keywords = "Category learning, Concepts, Error-driven learning, Unsupervised learning",
author = "Hoffman, {Aaron B.} and Murphy, {Gregory L.}",
year = "2006",
month = "3",
doi = "10.1037/0278-7393.32.3.301",
language = "English (US)",
volume = "32",
pages = "301--315",
journal = "Journal of Experimental Psychology: Learning Memory and Cognition",
issn = "0278-7393",
publisher = "American Psychological Association Inc.",
number = "2",

}

TY - JOUR

T1 - Category dimensionality and feature knowledge

T2 - When more features are learned as easily as fewer

AU - Hoffman, Aaron B.

AU - Murphy, Gregory L.

PY - 2006/3

Y1 - 2006/3

N2 - Three experiments compared the learning of lower-dimensional family resemblance categories (4 dimensions) with the learning of higher-dimensional ones (8 dimensions). Category-learning models incorporating error-driven learning, hypothesis testing, or limited capacity attention predict that additional dimensions should either increase learning difficulty or decrease learning of individual features. Contrary to these predictions, the experiments showed no slower learning of high-dimensional categories; instead, subjects learned more features from high-dimensional categories than from low-dimensional categories. This result obtained both in standard learning with feedback and in noncontingent, observational learning. These results show that rather than interfering with learning, categories with more dimensions cause individuals to learn more. The authors contrast the learning of family resemblance categories with learning in classical conditioning and probability learning paradigms, in which competition among features is well documented.

AB - Three experiments compared the learning of lower-dimensional family resemblance categories (4 dimensions) with the learning of higher-dimensional ones (8 dimensions). Category-learning models incorporating error-driven learning, hypothesis testing, or limited capacity attention predict that additional dimensions should either increase learning difficulty or decrease learning of individual features. Contrary to these predictions, the experiments showed no slower learning of high-dimensional categories; instead, subjects learned more features from high-dimensional categories than from low-dimensional categories. This result obtained both in standard learning with feedback and in noncontingent, observational learning. These results show that rather than interfering with learning, categories with more dimensions cause individuals to learn more. The authors contrast the learning of family resemblance categories with learning in classical conditioning and probability learning paradigms, in which competition among features is well documented.

KW - Category learning

KW - Concepts

KW - Error-driven learning

KW - Unsupervised learning

UR - http://www.scopus.com/inward/record.url?scp=33645727887&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33645727887&partnerID=8YFLogxK

U2 - 10.1037/0278-7393.32.3.301

DO - 10.1037/0278-7393.32.3.301

M3 - Article

VL - 32

SP - 301

EP - 315

JO - Journal of Experimental Psychology: Learning Memory and Cognition

JF - Journal of Experimental Psychology: Learning Memory and Cognition

SN - 0278-7393

IS - 2

ER -