en-de
en-es
en-fr
en-pt
en-sl
en
en-zh
0.25
0.5
0.75
1.25
1.5
1.75
2
Quasi-Random Resamplings, with Applications to Rule Extraction, Cross-Validation and (Su-)Bagging
Published on Feb 25, 20073413 Views
Chapter list
Quasi-random resampling00:00
What is the problem?00:17
A typical example01:09
Organization of the talk02:34
Why resampling is Monte-Carlo integration02:39
An introduction to QR-numbers04:30
QR-numbers04:53
Quasi-random numbers?05:31
Quasi-random = low discrepancy?06:39
A better discrepancy?07:48
Existing bounds on low-discrepancy-Monte-Carlo08:19
Which set do you trust?09:00
Which quasi-random numbers?09:45
What else than Monte-Carlo integration?10:21
... and how to do in strange spaces?10:49
Have fun with QR in strange spaces11:01
Quasi-random numbers in strange spaces11:31
... but we need something else!11:35
Quasi-random points in strange spaces11:57
Why not in something isotropic?12:05
For the gaussian: easy!12:11
Ok!13:22
However, we will do that14:19
QR-numbers in resampling pt 114:53
QR-numbers in resampling pt 215:18
QR-numbers in resampling pt 316:16
What are bootstrap samples?16:24
A naive solution pt 116:28
A naive solution pt 217:38
A naive solution pt 317:47
A naive solution pt 417:48
A naive solution pt 517:52
A naive solution pt 618:15
...which does not work18:58
A less naive solution19:08
With dimension-reduction it's better20:13
Summary21:07
Conclusions pt 121:11
Experiments21:14
A typical example24:22
Conclusions pt 225:42
Further work26:01