Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
On the utility of dreaming: A general model for how learning in artificial agents can benefit from data hallucination
Department of Computer Science, Middlesex University, UK / Centre for Vision, Speech and Signal Processing, University of Surrey, UK.
University of Skövde, School of Informatics. University of Skövde, The Informatics Research Centre. (Interaction Lab (ILAB))
University of Skövde, School of Informatics. University of Skövde, The Informatics Research Centre. Donders Institute for Brain, Cognition, and Behaviour, Radboud University, Netherlands. (Interaction Lab (ILAB))ORCID iD: 0000-0003-1177-4119
2020 (English)In: Adaptive Behavior, ISSN 1059-7123, E-ISSN 1741-2633, article id UNSP 1059712319896489Article in journal (Refereed) Epub ahead of print
Abstract [en]

We consider the benefits of dream mechanisms - that is, the ability to simulate new experiences based on past ones - in a machine learning context. Specifically, we are interested in learning for artificial agents that act in the world, and operationalize "dreaming" as a mechanism by which such an agent can use its own model of the learning environment to generate new hypotheses and training data. We first show that it is not necessarily a given that such a data-hallucination process is useful, since it can easily lead to a training set dominated by spurious imagined data until an ill-defined convergence point is reached. We then analyse a notably successful implementation of a machine learning-based dreaming mechanism by Ha and Schmidhuber (Ha, D., & Schmidhuber, J. (2018). World models. arXiv e-prints, arXiv:1803.10122). On that basis, we then develop a general framework by which an agent can generate simulated data to learn from in a manner that is beneficial to the agent. This, we argue, then forms a general method for an operationalized dream-like mechanism. We finish by demonstrating the general conditions under which such mechanisms can be useful in machine learning, wherein the implicit simulator inference and extrapolation involved in dreaming act without reinforcing inference error even when inference is incomplete.

Place, publisher, year, edition, pages
Sage Publications, 2020. article id UNSP 1059712319896489
Keywords [en]
Artificial dream mechanisms, data simulation, machine learning, reinforcement learning
National Category
Computer Sciences
Research subject
Interaction Lab (ILAB)
Identifiers
URN: urn:nbn:se:his:diva-18160DOI: 10.1177/1059712319896489ISI: 000506780000001Scopus ID: 2-s2.0-85077601986OAI: oai:DiVA.org:his-18160DiVA, id: diva2:1388030
Available from: 2020-01-23 Created: 2020-01-23 Last updated: 2020-04-02Bibliographically approved

Open Access in DiVA

fulltext(767 kB)16 downloads
File information
File name FULLTEXT01.pdfFile size 767 kBChecksum SHA-512
a7e794105303b04ee3057ae5616def4b5cf14fe2eefcc3bbc05320d916582dc28e01b06794118ffa2381595c2926a99cc4906d31a745586e6c0218df062d11d5
Type fulltextMimetype application/pdf

Other links

Publisher's full textScopus

Search in DiVA

By author/editor
Svensson, HenrikThill, Serge
By organisation
School of InformaticsThe Informatics Research Centre
In the same journal
Adaptive Behavior
Computer Sciences

Search outside of DiVA

GoogleGoogle Scholar
Total: 16 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

doi
urn-nbn

Altmetric score

doi
urn-nbn
Total: 49 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf