Publications

BibTex format

@inproceedings{Shanahan:2022:ijcai.2022/780,
author = {Shanahan, M and Mitchell, M},
doi = {ijcai.2022/780},
pages = {5588--5596},
publisher = {IJCAI},
title = {Abstraction for deep reinforcement learning},
url = {http://dx.doi.org/10.24963/ijcai.2022/780},
year = {2022}
}

Download

RIS format (EndNote, RefMan)

TY  - CPAPER
AB  - We characterise the problem of abstraction in thecontext of deep reinforcement learning. Variouswell established approaches to analogical reasoningand associative memory might be brought to bearon this issue, but they present difficulties becauseof the need for end-to-end differentiability. We re-view developments in AI and machine learning thatcould facilitate their adoption.
AU  - Shanahan,M
AU  - Mitchell,M
DO  - ijcai.2022/780
EP  - 5596
PB  - IJCAI
PY  - 2022///
SN  - 1045-0823
SP  - 5588
TI  - Abstraction for deep reinforcement learning
UR  - http://dx.doi.org/10.24963/ijcai.2022/780
UR  - https://www.ijcai.org/proceedings/2022/780
UR  - http://hdl.handle.net/10044/1/97606
ER  -

Download

ProfessorMurrayShanahan

Contact

Location

Summary

Citation

BibTex format

RIS format (EndNote, RefMan)