Modeling unsupervised perceptual category learning

Brenden M. Lake, Gautam K. Vallabha, James L. McClelland

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

During the learning of speech sounds and other perceptual categories, category labels are not provided, the number of categories is unknown, and the stimuli are encountered sequentially. These constraints provide a challenge for models, but they have been recently addressed in the Online Mixture Estimation model of unsupervised vowel category learning [1]. the model treats categories as Gaussian distributions, proposing both the number and parameters of the categories. while the model has been shown to successfully learn vowel categories, it has not been evaluated as a model of the learning process. We account for three results regarding the learning process: Infants' discrimination of speech sounds is better after exposure to a bimodal rather than unimodal distribution [2], infants' discrimination of vowels is affected by acoustic distance [3], and subjects place category centers near frequent stimuli in an unsupervised visual classification task [4].

Original languageEnglish (US)
Title of host publication2008 IEEE 7th International Conference on Development and Learning, ICDL
Pages25-30
Number of pages6
DOIs
StatePublished - 2008
Event2008 IEEE 7th International Conference on Development and Learning, ICDL - Monterey, CA, United States
Duration: Aug 9 2008Aug 12 2008

Publication series

Name2008 IEEE 7th International Conference on Development and Learning, ICDL

Other

Other2008 IEEE 7th International Conference on Development and Learning, ICDL
Country/TerritoryUnited States
CityMonterey, CA
Period8/9/088/12/08

ASJC Scopus subject areas

  • Developmental Biology
  • General Computer Science
  • Biomedical Engineering
  • Education

Fingerprint

Dive into the research topics of 'Modeling unsupervised perceptual category learning'. Together they form a unique fingerprint.

Cite this