Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Neural nets for generalization and classification: comment on Staddon and Reid (1990)

R N Shepard1

  • 1Department of Psychology, Stanford University, California 94305-2130.

Psychological Review
|October 1, 1990
PubMed
Summary

Neural network models explain generalization gradients similarly to diffusion models. Cognitive generalization theory, implemented in connectionist networks, offers a more comprehensive account of classification learning.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

The time required to prepare for a rotated stimulus.

Memory & cognition·2013
Same author

Perceptual-cognitive universals as reflections of the world.

Psychonomic bulletin & review·2013
Same author

Undirected motility of filamentous cyanobacteria produces reticulate mats.

Geobiology·2010
Same author

Perceptual-cognitive universals as reflections of the world.

The Behavioral and brain sciences·2002
Same author

Carl Iver Hovland: June 12, 1912-April 16, 1961.

Biographical memoirs. National Academy of Sciences (U.S.)·2001
Same author

Quantitation of human immunodeficiency virus type 1 RNA in different biological compartments.

Journal of clinical microbiology·2000

Area of Science:

  • Cognitive Science
  • Computational Neuroscience
  • Machine Learning Theory

Background:

  • The neural network model by Staddon and Reid (1990) provides a framework for understanding generalization gradients.
  • Shepard's (1958) diffusion model offers an early explanation for exponential and Gaussian generalization patterns.

Discussion:

  • This study compares the explanatory power of Staddon and Reid's neural network model with Shepard's diffusion model regarding generalization gradients.
  • It highlights how both models account for similar phenomena in generalization.
  • The cognitive generalization theory, also implemented as a connectionist network, is discussed as an advancement.

Key Insights:

  • Neural network models and diffusion models offer converging explanations for exponential and Gaussian generalization gradients.

Related Experiment Videos

  • Shepard's cognitive generalization theory, when implemented in a connectionist network, provides a richer account of classification learning.
  • Connectionist implementations of cognitive theories offer enhanced explanatory power.

    • Outlook:

    • Future research could explore further integration of cognitive theories within connectionist architectures.
    • Investigating the application of these models to diverse learning tasks could yield new insights.
    • Further computational modeling may refine our understanding of generalization mechanisms.