Jove
Visualize
Contact Us

Related Experiment Videos

Key Concepts in Model Selection: Performance and Generalizability.

Forster1

  • 1University of Wisconsin, Madison

Journal of Mathematical Psychology
|March 29, 2000
PubMed
Summary
This summary is machine-generated.

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

Evidence for a Soft Nuclear Equation-of-State from Kaon Production in Heavy-Ion Collisions.

Physical review letters·2001
Same author

Monte-Carlo-based investigations of the edge effect of endovascular brachytherapy sources.

Cardiovascular radiation medicine·2000
Same author

Present and future capabilities of MCNP

Applied radiation and isotopes : including data, instrumentation and methods for use in agriculture, industry and medicine·2000
Same author

Yield optimization and time structure of femtosecond laser plasma kalpha sources

Physical review letters·2000
Same author

The steam volatile oil of Wollemia nobilis and its comparison with other members of the Araucariaceae (Agathis and Araucaria).

Biochemical systematics and ecology·2000
Same author

GUEST EDITORS' INTRODUCTION.

Journal of mathematical psychology·2000
Same journal

Corrigendum to "An entropy model of decision uncertainty" [Journal of Mathematical Psychology 125 (2025), 102919].

Journal of mathematical psychology·2025
Same journal

An entropy model of decision uncertainty.

Journal of mathematical psychology·2025
Same journal

How do people build up visual memory representations from sensory evidence? Revisiting two classic models of choice.

Journal of mathematical psychology·2024
Same journal

Experiment-based calibration in psychology: Optimal design considerations.

Journal of mathematical psychology·2024
Same journal

Expressions for Bayesian confidence of drift diffusion observers in fluctuating stimuli tasks.

Journal of mathematical psychology·2024
Same journal

A Statistical Foundation for Derived Attention.

Journal of mathematical psychology·2023
See all related articles
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

Model selection balances simplicity with data fit, but standard methods may not fully address generalization errors. Further research is needed to improve model selection for broader predictive accuracy.

Area of Science:

  • Statistical modeling
  • Machine learning theory

Background:

  • Model selection is crucial for choosing appropriate statistical models.
  • Existing methods often focus on parameter estimation errors.

Purpose of the Study:

  • To explain fundamental concepts in model selection.
  • To highlight existing research and introduce novel perspectives.
  • To address the limitations of current model selection techniques.

Main Methods:

  • Classical hypothesis testing
  • Maximum likelihood estimation
  • Bayesian methods
  • Minimum Description Length (MDL)
  • Cross-validation
  • Akaike's Information Criterion (AIC)

Related Experiment Videos

Main Results:

  • Standard methods implement Occam's razor, balancing parsimony and goodness-of-fit.
  • These methods primarily address sampling errors in parameter estimation.
  • Generalization errors (extrapolation) are distinct from parameter estimation errors.

Conclusions:

  • Current model selection methods are incomplete implementations of Occam's razor.
  • Model selection should explicitly account for generalization ability.
  • Simplicity and parsimony play a role in managing extrapolation errors.