Principled methods of trading exploration and exploitation

Models for Trading Exploration and Exploitation using Upper Confidence Bounds

author: Peter Auer, University of Leoben

You might be experiencing some problems with Your Video player.

*Slides*
0:00	Models for Trading
0:39	Overview
1:25	The bandit problem with linear side information
3:08	Goal
4:42	Relation to other models
6:54	Results
9:16	Remarks
9:53	Algorithm: Using upper confidence bounds
11:28	Why does this work?
12:58	Algorithm: Using upper confidence bounds
13:12	Why does this work?
13:35	Algorithm: Using upper confidence bounds
15:40	Why does this work?
16:05	Bounding the widths of the confidence intervals
17:55	Random bandit problem (cont.)
20:30	Random bandit problem (cont.)
22:13	Random bandit problem: Improved bounds
23:08	Random bandit problem: Improved bounds
23:20	Random bandit problem: Improved bounds
23:41	Random bandit problem: Improved bounds
23:47	Random bandit problem: Improved bounds
24:10	Random bandit problem: Improved bounds
24:29	Random bandit problem: Improved bounds
26:20	Random bandit problem (cont.)
27:26	Linear side information
31:25	Calculating the variance
32:21	Optimizing the regret
33:20	Linear side information
33:58	Optimizing the regret
34:20	Calculating
35:50	Optimizing the regret
35:58	Calculating
37:41	Reinforcement learning
37:55	Calculating
40:09	Reinforcement learning
43:05	Motivation for online reinforcement learning
44:31	Discounted and undiscounted returns
46:48	Regret
47:41	Episodic reinforcement learning
49:18	The algorithm: Upper confidence bounds again
53:17	Why it works (1)
55:07	Why it works (2)
56:04	Why it works (3)
59:46	Conclusion
62:28	auer-slides_Page_34

Lecture rating

People found this lecture:
Worth seeing
because it is:
Valuable and informative
Well presented
Easily understandable
Acceptably recorded
You need to login to cast your vote.

Report a problem or upload files

If you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.