Advanced search
1 file | 328.88 KB Add to list

A learning gap between neuroscience and reinforcement learning

Samuel Wauthier (UGent) , Pietro Mazzaglia (UGent) , Ozan Catal (UGent) , Cedric De Boom (UGent) , Tim Verbelen (UGent) and Bart Dhoedt (UGent)
Author
Organization
Project
Abstract
Historically, artificial intelligence has drawn much inspiration from neuroscience to fuel advances in the field. However, current progress in reinforcement learning is largely focused on benchmark problems that fail to capture many of the aspects that are of interest in neuroscience today. We illustrate this point by extending a T-maze task from neuroscience for use with reinforcement learning algorithms, and show that state-of-the-art algorithms are not capable of solving this problem. Finally, we point out where insights from neuroscience could help explain some of the issues encountered.

Downloads

  • 7924.pdf
    • full text (Accepted manuscript)
    • |
    • open access
    • |
    • PDF
    • |
    • 328.88 KB

Citation

Please use this url to cite or link to this publication:

MLA
Wauthier, Samuel, et al. “A Learning Gap between Neuroscience and Reinforcement Learning.” BRAIN2AI, How Can Findings About The Brain Improve AI Systems, ICLR 2021 Workshop, Proceedings, 2021.
APA
Wauthier, S., Mazzaglia, P., Catal, O., De Boom, C., Verbelen, T., & Dhoedt, B. (2021). A learning gap between neuroscience and reinforcement learning. In BRAIN2AI, How Can Findings About The Brain Improve AI Systems, ICLR 2021 Workshop, Proceedings. Online.
Chicago author-date
Wauthier, Samuel, Pietro Mazzaglia, Ozan Catal, Cedric De Boom, Tim Verbelen, and Bart Dhoedt. 2021. “A Learning Gap between Neuroscience and Reinforcement Learning.” In BRAIN2AI, How Can Findings About The Brain Improve AI Systems, ICLR 2021 Workshop, Proceedings.
Chicago author-date (all authors)
Wauthier, Samuel, Pietro Mazzaglia, Ozan Catal, Cedric De Boom, Tim Verbelen, and Bart Dhoedt. 2021. “A Learning Gap between Neuroscience and Reinforcement Learning.” In BRAIN2AI, How Can Findings About The Brain Improve AI Systems, ICLR 2021 Workshop, Proceedings.
Vancouver
1.
Wauthier S, Mazzaglia P, Catal O, De Boom C, Verbelen T, Dhoedt B. A learning gap between neuroscience and reinforcement learning. In: BRAIN2AI, How Can Findings About The Brain Improve AI Systems, ICLR 2021 Workshop, Proceedings. 2021.
IEEE
[1]
S. Wauthier, P. Mazzaglia, O. Catal, C. De Boom, T. Verbelen, and B. Dhoedt, “A learning gap between neuroscience and reinforcement learning,” in BRAIN2AI, How Can Findings About The Brain Improve AI Systems, ICLR 2021 Workshop, Proceedings, Online, 2021.
@inproceedings{8712808,
  abstract     = {{Historically, artificial intelligence has drawn much inspiration from
neuroscience to fuel advances in the field. However, current progress in
reinforcement learning is largely focused on benchmark problems that fail to
capture many of the aspects that are of interest in neuroscience today. We
illustrate this point by extending a T-maze task from neuroscience for use with
reinforcement learning algorithms, and show that state-of-the-art algorithms
are not capable of solving this problem. Finally, we point out where insights
from neuroscience could help explain some of the issues encountered.}},
  author       = {{Wauthier, Samuel and Mazzaglia, Pietro and Catal, Ozan and De Boom, Cedric and Verbelen, Tim and Dhoedt, Bart}},
  booktitle    = {{BRAIN2AI, How Can Findings About The Brain Improve AI Systems, ICLR 2021 Workshop, Proceedings}},
  language     = {{eng}},
  location     = {{Online}},
  pages        = {{9}},
  title        = {{A learning gap between neuroscience and reinforcement learning}},
  url          = {{https://iclrbrain2ai.github.io/#call-for-papers}},
  year         = {{2021}},
}