About | Contact Us | Register | Login
ProceedingsSeriesJournalsSearchEAI
Intelligent Technologies for Interactive Entertainment. 12th EAI International Conference, INTETAIN 2020, Virtual Event, December 12-14, 2020, Proceedings

Research Article

Neuroevolution vs Reinforcement Learning for Training Non Player Characters in Games: The Case of a Self Driving Car

Download(Requires a free EAI acccount)
5 downloads
Cite
BibTeX Plain Text
  • @INPROCEEDINGS{10.1007/978-3-030-76426-5_13,
        author={Kristi\^{a}n Kovalsk\"{y} and George Palamas},
        title={Neuroevolution vs Reinforcement Learning for Training Non Player Characters in Games: The Case of a Self Driving Car},
        proceedings={Intelligent Technologies for Interactive Entertainment. 12th EAI International Conference, INTETAIN 2020, Virtual Event, December 12-14, 2020, Proceedings},
        proceedings_a={INTETAIN},
        year={2021},
        month={5},
        keywords={Neuroevolution Reinforcement learning Neural network Evolutionary algorithm Autonomous systems Self driving car Unity Games Non player character NPC},
        doi={10.1007/978-3-030-76426-5_13}
    }
    
  • Kristián Kovalský
    George Palamas
    Year: 2021
    Neuroevolution vs Reinforcement Learning for Training Non Player Characters in Games: The Case of a Self Driving Car
    INTETAIN
    Springer
    DOI: 10.1007/978-3-030-76426-5_13
Kristián Kovalský1,*, George Palamas1
  • 1: Aalborg University, A. C. Meyers Vænge 15
*Contact email: aau@aau.dk

Abstract

The aim of this project is to compare two popular machine learning methods, a non-gradient-based algorithm such as neuro-evolution with a gradient-based reinforcement learning on an irregular task of training a car to self-drive around 3D circuits with varying complexity. A series of 3D circuits with a physics based car model were modeled using the Unity game engine. The data collected during evaluation show that neuro-evolution converges faster to a solution when compared to the reinforcement learning approach. However, when the reinforcement learning approach is allowed to train for long enough, it outperforms the neuro-evolution in terms of car speed and lap times achieved by the trained model of the car.

Keywords
Neuroevolution Reinforcement learning Neural network Evolutionary algorithm Autonomous systems Self driving car Unity Games Non player character NPC
Published
2021-05-19
Appears in
SpringerLink
http://dx.doi.org/10.1007/978-3-030-76426-5_13
Copyright © 2020–2025 ICST
EBSCOProQuestDBLPDOAJPortico
EAI Logo

About EAI

  • Who We Are
  • Leadership
  • Research Areas
  • Partners
  • Media Center

Community

  • Membership
  • Conference
  • Recognition
  • Sponsor Us

Publish with EAI

  • Publishing
  • Journals
  • Proceedings
  • Books
  • EUDL