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

Learning tactical human behavior through observation of human performance.

Hans K G Fernlund1, Avelino J Gonzalez, Michael Georgiopoulos

  • 1Department of Culture, Media and Computer Sciences, Dalarna University, Dalarna, 78188 Borlänge, Sweden. hfe@du.se

IEEE Transactions on Systems, Man, and Cybernetics. Part B, Cybernetics : a Publication of the IEEE Systems, Man, and Cybernetics Society
|February 14, 2006
PubMed
Summary

This study introduces a new method using genetic programming and context-based reasoning to automatically create simulated tactical agents by observing human performance. The evolved agents accurately mimic and generalize human behavior in simulations.

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

Autonomous Binarized Focal Loss Enhanced Model Compression Design Using Tensor Train Decomposition.

Micromachines·2022
Same author

Pareto-Optimal Model Selection via SPRINT-Race.

IEEE transactions on cybernetics·2017
Same author

Behavioral Modeling Based on Probabilistic Finite Automata: An Empirical Study.

Sensors (Basel, Switzerland)·2016
Same author

Multi-Objective Model Selection via Racing.

IEEE transactions on cybernetics·2015
Same author

A Simple Method for Solving the SVM Regularization Path for Semidefinite Kernels.

IEEE transactions on neural networks and learning systems·2015
Same author

Pareto-path multitask multiple kernel learning.

IEEE transactions on neural networks and learning systems·2014

Area of Science:

  • Artificial Intelligence
  • Computer Science
  • Simulation Modeling

Background:

  • Acquiring tactical knowledge for simulated agents is costly and difficult.
  • Eliciting and representing expert knowledge through traditional methods is challenging.
  • Existing approaches have had limited success in creating realistic tactical agents.

Purpose of the Study:

  • To develop a novel approach for automatically evolving tactical agents.
  • To overcome the limitations of knowledge acquisition in simulation development.
  • To create agents that accurately reflect human tactical performance.

Main Methods:

  • Employing genetic programming (GP) combined with context-based reasoning.
  • Utilizing automatic observation of human performance in a mission simulator.

Related Experiment Videos

  • Developing a prototype system to demonstrate the feasibility of the approach.

    • Main Results:

    • Evolved agents demonstrated high fidelity to observed human performance.
    • The agents showed a capacity to generalize learned tactical behaviors.
    • The prototype system proved the viability of the automated learning process.

    Conclusions:

    • The GP and context-based reasoning approach effectively automates tactical agent development.
    • This method significantly reduces the cost and difficulty of creating realistic simulated agents.
    • The evolved agents offer a promising solution for military and game simulations.