Meta-learning for deep reinforcement learning
Creating a fully autonomous agent that is able to solve real world tasks at superhuman level is always one of the primary goals in the field of artificial intelligence (AI). Most works that are done in AI field until today are categorized as ”narrow” AI, even though the field of AI has been in th...
Saved in:
Main Author: | |
---|---|
Other Authors: | |
Format: | Final Year Project |
Language: | English |
Published: |
Nanyang Technological University
2021
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/148139 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
id |
sg-ntu-dr.10356-148139 |
---|---|
record_format |
dspace |
spelling |
sg-ntu-dr.10356-1481392021-04-24T04:31:13Z Meta-learning for deep reinforcement learning Poon, Jun Yaw Sinno Jialin Pan School of Computer Science and Engineering Computational Intelligence Lab sinnopan@ntu.edu.sg Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence Creating a fully autonomous agent that is able to solve real world tasks at superhuman level is always one of the primary goals in the field of artificial intelligence (AI). Most works that are done in AI field until today are categorized as ”narrow” AI, even though the field of AI has been in the world for seventy years [1]. A more ambitious goal in AI is to create Artificial General Intelligence (AGI). The concept of meta-learning is one step closer to this ambitious goal of AI. In this project, we studied existing meta-learning algorithms for deep reinforcement learning, namely MAML and PEARL, and compare them in terms of sample efficiency and their performance in terms of the final average returns. We also studied an existing algorithm in RL, namely SLAC, that addresses the latent representation problem in RL by learning a sequential latent variable model. We integrated SLAC into PEARL and compare the its performance to the original PEARL. Our work shows that the integrated variant of PEARL with SLAC has a better performance than the original PEARL. Bachelor of Engineering (Computer Engineering) 2021-04-24T04:31:13Z 2021-04-24T04:31:13Z 2021 Final Year Project (FYP) Poon, J. Y. (2021). Meta-learning for deep reinforcement learning. Final Year Project (FYP), Nanyang Technological University, Singapore. https://hdl.handle.net/10356/148139 https://hdl.handle.net/10356/148139 en SCSE20-0441 application/pdf Nanyang Technological University |
institution |
Nanyang Technological University |
building |
NTU Library |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
NTU Library |
collection |
DR-NTU |
language |
English |
topic |
Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence |
spellingShingle |
Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence Poon, Jun Yaw Meta-learning for deep reinforcement learning |
description |
Creating a fully autonomous agent that is able to solve real world tasks at superhuman level
is always one of the primary goals in the field of artificial intelligence (AI). Most works that
are done in AI field until today are categorized as ”narrow” AI, even though the field of AI
has been in the world for seventy years [1]. A more ambitious goal in AI is to create Artificial
General Intelligence (AGI). The concept of meta-learning is one step closer to this ambitious
goal of AI. In this project, we studied existing meta-learning algorithms for deep reinforcement
learning, namely MAML and PEARL, and compare them in terms of sample efficiency and
their performance in terms of the final average returns. We also studied an existing algorithm
in RL, namely SLAC, that addresses the latent representation problem in RL by learning
a sequential latent variable model. We integrated SLAC into PEARL and compare the its
performance to the original PEARL. Our work shows that the integrated variant of PEARL
with SLAC has a better performance than the original PEARL. |
author2 |
Sinno Jialin Pan |
author_facet |
Sinno Jialin Pan Poon, Jun Yaw |
format |
Final Year Project |
author |
Poon, Jun Yaw |
author_sort |
Poon, Jun Yaw |
title |
Meta-learning for deep reinforcement learning |
title_short |
Meta-learning for deep reinforcement learning |
title_full |
Meta-learning for deep reinforcement learning |
title_fullStr |
Meta-learning for deep reinforcement learning |
title_full_unstemmed |
Meta-learning for deep reinforcement learning |
title_sort |
meta-learning for deep reinforcement learning |
publisher |
Nanyang Technological University |
publishDate |
2021 |
url |
https://hdl.handle.net/10356/148139 |
_version_ |
1698713640967864320 |