event thumbnail image
International Workshop on Intelligent Information Access

Quasi-Random Resamplings, with Applications to Rule Extraction, Cross-Validation and (Su-)Bagging

author: Olivier Teytaud, INRIA
You might be experiencing some problems with Your Video player.
Slides
0:00 Quasi-random resampling
0:17 What is the problem?
1:09 A typical example
2:34 Organization of the talk
2:39 Why resampling is Monte-Carlo integration
4:30 An introduction to QR-numbers
4:53 QR-numbers
5:31 Quasi-random numbers?
6:39 Quasi-random = low discrepancy?
7:48 A better discrepancy?
8:19 Existing bounds on low-discrepancy-Monte-Carlo
9:00 Which set do you trust?
9:45 Which quasi-random numbers?
10:21 What else than Monte-Carlo integration?
10:49 ... and how to do in strange spaces?
11:01 Have fun with QR in strange spaces
11:31 Quasi-random numbers in strange spaces
11:35 ... but we need something else!
11:57 Quasi-random points in strange spaces
12:05 Why not in something isotropic?
12:11 For the gaussian: easy!
13:22 Ok!
14:19 However, we will do that
14:53 QR-numbers in resampling pt 1
15:18 QR-numbers in resampling pt 2
16:16 QR-numbers in resampling pt 3
16:24 What are bootstrap samples?
16:28 A naive solution pt 1
17:38 A naive solution pt 2
17:47 A naive solution pt 3
17:48 A naive solution pt 4
17:52 A naive solution pt 5
18:15 A naive solution pt 6
18:58 ...which does not work
19:08 A less naive solution
20:13 With dimension-reduction it's better
21:07 Summary
21:11 Conclusions pt 1
21:14 Experiments
24:22 A typical example
25:42 Conclusions pt 2
26:01 Further work

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.

Link this page

Would you like to put a link to this lecture on your homepage?
Go ahead! Copy the HTML snippet !

Write your own review or comment:

make sure you have javascript enabled or clear this field: