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

Failure of maximum likelihood methods for chaotic dynamical systems.

Kevin Judd1

  • 1School of Mathematics and Statistics, University of Western Australia, Perth, Australia.

Physical Review. E, Statistical, Nonlinear, and Soft Matter Physics
|May 16, 2007
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

Growing optimal scale-free networks via likelihood.

Physical review. E, Statistical, nonlinear, and soft matter physics·2015
Same author

Networked dynamical systems with linear coupling: synchronisation patterns, coherence and other behaviours.

Chaos (Woodbury, N.Y.)·2014
Same author

Forecasting: it is not about statistics, it is about dynamics.

Philosophical transactions. Series A, Mathematical, physical, and engineering sciences·2009
Same author

Failures of sequential Bayesian filters and the successes of shadowing filters in tracking of nonlinear deterministic and stochastic systems.

Physical review. E, Statistical, nonlinear, and soft matter physics·2009
Same author

The seasonal relationship between assault and homicide in England and Wales.

Injury·2008
Same author

Degeneracy of time series models: the best model is not always the correct model.

Chaos (Woodbury, N.Y.)·2006

Maximum likelihood estimation fails for chaotic dynamical systems, as alternative trajectories can appear infinitely more likely. This impacts parameter estimation, suggesting alternative methods are needed for accurate analysis.

Area of Science:

  • Statistics
  • Dynamical Systems Theory
  • Chaos Theory

Background:

  • Maximum likelihood (ML) is a fundamental statistical technique for parameter estimation.
  • It serves as a basis for more advanced methods, including Bayesian approaches.
  • Understanding the limitations of ML is crucial for robust data analysis.

Purpose of the Study:

  • To investigate the efficacy of maximum likelihood in identifying the true trajectory of chaotic dynamical systems.
  • To determine the conditions under which maximum likelihood estimation fails in these systems.
  • To propose potential remedies for the identified failures.

Main Methods:

  • Theoretical analysis of maximum likelihood estimation applied to chaotic dynamical systems.
  • Examination of likelihood calculations for true versus alternative trajectories.

Related Experiment Videos

  • Simulation-based assessment of failure conditions with varying noise types and magnitudes.

    • Main Results:

    • Maximum likelihood fails to identify the true trajectory in chaotic systems due to the existence of "truth-dominating" trajectories.
    • This failure occurs with both unbounded noise and sufficiently large bounded noise.
    • The residuals of these dominating trajectories exhibit characteristics inconsistent with the assumed noise distribution.

    Conclusions:

    • Maximum likelihood estimation is unreliable for parameter estimation in chaotic dynamical systems under certain noise conditions.
    • Alternative statistical methods that analyze residual distributions are recommended.
    • Future research should focus on developing robust estimation techniques for chaotic systems.