en-es
en
0.25
0.5
0.75
1.25
1.5
1.75
2
Information Complexity in Bandit Subset Selection
Published on Aug 09, 20132880 Views
We consider the problem of efficiently exploring the arms of a stochastic bandit to identify the best subset. Under the PAC and the fixed-budget formulations, we derive improved bounds by using KL-div
Related categories
Chapter list
Information complexity in bandit subset selection00:00
Stochastic Bandit: One statistical model00:11
Stochastic Bandit: Two objectives - 100:54
Stochastic Bandit: Two objectives - 201:20
The Explore-m problem - 101:43
The Explore-m problem - 202:11
The Explore-m problem - 302:33
Open questions for Explore-m - 102:44
Open questions for Explore-m - 203:20
Our contributions - 103:46
Our contributions - 204:09
Our contributions - 304:21
KL-Racing: uniform sampling and eliminations04:33
KL-LUCB: adaptive sampling05:18
Sample Complexity bound for KL-LUCB05:45