Application of Newton's method to action selection in continuous state- and action-space reinforcement learning

Conference paper


Nichols, B. and Dracopoulos, D. 2014. Application of Newton's method to action selection in continuous state- and action-space reinforcement learning. The European Symposium on Artificial Neural Networks (ESANN). ESANN. pp. 141-146
TypeConference paper
TitleApplication of Newton's method to action selection in continuous state- and action-space reinforcement learning
AuthorsNichols, B. and Dracopoulos, D.
Abstract

An algorithm based on Newton’s Method is proposed for action selection in continuous state- and action-space reinforcement learning without a policy network or discretization. The proposed method is validated on two benchmark problems: Cart-Pole and double Cart-Pole on which the proposed method achieves comparable or improved performance with less parameters to tune and in less training episodes than CACLA, which has previously been shown to outperform many other continuous state- and action-space reinforcement learning algorithms.

ConferenceThe European Symposium on Artificial Neural Networks (ESANN)
Page range141-146
ISBN
Hardcover9782874190957
PublisherESANN
Publication dates
Print17 Mar 2014
Publication process dates
Deposited24 Jun 2015
Output statusPublished
Publisher's version
Web address (URL)https://www.elen.ucl.ac.be/Proceedings/esann/esannpdf/es2014-175.pdf
LanguageEnglish
Book titleESANN 2014 Proceedings: 22nd European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning. Bruges April 23-24-25, 2014
Permalink -

https://repository.mdx.ac.uk/item/85q37

Download files


Publisher's version
  • 14
    total views
  • 1
    total downloads
  • 2
    views this month
  • 0
    downloads this month

Export as