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

Modeling and prediction of human behavior

A Pentland1, A Liu

  • 1The Media Laboratory, E15-387, Massachusetts Institute of Technology, 20 Ames Street, Cambridge MA 02139, USA. sandy@media.mit.edu

Neural Computation
|February 9, 1999
PubMed
Summary

This study models human behavior using dynamic modes sequenced by Markov chains. The approach achieved 95% accuracy in predicting drivers' actions from preparatory movements.

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

Black Hole Spectroscopy and Tests of General Relativity with GW250114.

Physical review letters·2026
Same author

The Ruminant Farm Systems (RuFaS) model is a platform to support future research and actions for sustainable dairy farming.

JDS communications·2026
Same author

Radiomic signatures to estimate survival in patients with advanced hepatocellular carcinoma treated with sorafenib: Cancer and Leukemia Group B 80802 (Alliance).

ESMO open·2025
Same author

Adverse events associated with sequential immune checkpoint inhibitor and alectinib in patients with ALK-rearranged advanced non-small-cell lung cancer.

ESMO open·2025
Same author

GW250114: Testing Hawking's Area Law and the Kerr Nature of Black Holes.

Physical review letters·2025
Same author

Unifying measurement schemes in 2D terahertz spectroscopy.

The Journal of chemical physics·2025

Area of Science:

  • Computational neuroscience
  • Machine learning
  • Behavioral modeling

Background:

  • Human behavior is complex and challenging to model computationally.
  • Existing models may not fully capture the dynamic and sequential nature of actions.

Purpose of the Study:

  • To propose and validate a novel framework for modeling human behavior using dynamic Markov models.
  • To demonstrate the efficacy of this approach in recognizing and predicting human actions from sensory data.

Main Methods:

  • Developed dynamic Markov models combining dynamic modes (e.g., Kalman filters) with Markov chains.
  • Applied these models to recognize and predict human behaviors from sensory input.
  • Conducted an experiment involving automobile drivers to test prediction accuracy.

Main Results:

  • Achieved 95% accuracy in predicting automobile drivers' subsequent actions.
  • Demonstrated successful recognition of human behaviors from sensory data.
  • Validated the effectiveness of the dynamic Markov model approach.

Conclusions:

  • Dynamic Markov models provide an accurate and powerful method for describing and predicting human behavior.
  • This framework has significant potential for applications in human-computer interaction, robotics, and safety systems.
  • The model's success in predicting driver actions highlights its real-world applicability.

Related Experiment Videos