[Sds-seminars] S&DS Seminar, Speaker: Wen Sun, 9/12 @ 4pm-5pm, "Efficient Rich-observation Reinforcement Learning: A Representation Learning Approach"

[elizavette.torres at yale.edu]

   <https://statistics.yale.edu/> Department of Statistics and Data Science


 

WEN SUN,  Cornell University

Date: Monday, Sept. 12, 2022

Time: 4:00PM to 5:00PM

Zoom:
<https://yale.zoom.us/j/92411077917?pwd=aXhnTnFGRXFoaTVDczNjeFFKeWpTQT09>
https://yale.zoom.us/j/92411077917?pwd=aXhnTnFGRXFoaTVDczNjeFFKeWpTQT09,
Password: 24

 

 

 

Title: Efficient Rich-observation Reinforcement Learning: A Representation
Learning Approach

 

Abstract: Representation learning is a promising approach for solving
large-scale Reinforcement Learning (RL) problems where data is complex and
high-dimensional (i.e., rich-observation RL). While representation learning
in computer vision and natural language processing has made significant
empirical progress, the understanding of its application to RL is still
limited, due to the unique challenges from RL such as the interplay between
representation learning and exploration (e.g., deep exploration requires
informative representation which in turn cannot be easily discovered without
adequate exploration). In this talk, we study representation learning for RL
in the context of low-rank MDPs where features are unknown a priori, which
forces us to think about the intricate coupling of representation learning
and decision making. We develop two styles of representation learning
methods, one is based on model learning, and the other one is based
adversarial training. Both methods interleave representation learning,
exploration, and exploitation to achieve polynomial sample complexity in
learning near optimal policies. Empirically, we demonstrate our approach on
a set of challenging rich observation RL benchmarks which require deep
exploration, and show that our approach outperforms prior state-of-art
theoretical approaches and also empirical deep RL baselines.

 

This is joint work with Masa Uehara, Xuezhou Zhang, Yuda Song, Mengdi Wang,
and Alekh Agarwal.

 

Via Zoom: Join from PC, Mac, Linux, iOS or Android:
<https://yale.zoom.us/j/92411077917?pwd=aXhnTnFGRXFoaTVDczNjeFFKeWpTQT09>
https://yale.zoom.us/j/92411077917?pwd=aXhnTnFGRXFoaTVDczNjeFFKeWpTQT09

    Password: 24

    Or Telephone：203-432-9666 (2-ZOOM if on-campus) or 646 568 7788

    Meeting ID: 924 1107 7917

    International numbers available:  <https://yale.zoom.us/u/ad1lmXjRIM>
https://yale.zoom.us/u/ad1lmXjRIM

For H.323 and SIP information for video conferencing units please click
here:
<https://yale.service-now.com/it?id=support_article&sys_id=434b72d3db9e8fc83
514b1c0ef961924>
https://yale.service-now.com/it?id=support_article&sys_id=434b72d3db9e8fc835
14b1c0ef961924

 

3:45pm -4:00pm - meet and greet

4:00pm-5:00pm - Talk

For more details and upcoming events visit our website at
<http://statistics.yale.edu/> http://statistics.yale.edu/ .

 

 

 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.yale.edu/pipermail/sds-seminars/attachments/20220907/599b76a8/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.png
Type: image/png
Size: 6496 bytes
Desc: not available
URL: <http://mailman.yale.edu/pipermail/sds-seminars/attachments/20220907/599b76a8/attachment.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image004.jpg
Type: image/jpeg
Size: 15994 bytes
Desc: not available
URL: <http://mailman.yale.edu/pipermail/sds-seminars/attachments/20220907/599b76a8/attachment.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image002.jpg
Type: image/jpeg
Size: 2410 bytes
Desc: not available
URL: <http://mailman.yale.edu/pipermail/sds-seminars/attachments/20220907/599b76a8/attachment-0001.jpg>