0.25
0.5
0.75
1.25
1.5
1.75
2
A selecting-the-best method for budgeted model selection
Published on Nov 30, 20112505 Views
The paper focuses on budgeted model selection, that is the selection between a set of alternative models when the ratio between the number of model assessments and the number of alternatives, though b
Related categories
Chapter list
A selecting-the-best method for budgeted model selection00:00
Context and motivation00:12
Selection in a stochastic setting01:42
Multi-armed bandit and selecting-the-best03:01
Contribution04:35
Formal setting - 105:34
Formal setting - 206:32
The K = 2 Gaussian case - 107:13
The K = 2 Gaussian case - 208:14
The K = 2 Gaussian case - 309:36
The K = 2 optimal sampling strategy10:50
The K > 2 case11:35
Clark approximation12:35
K > 2 sampling strategy13:25
The K > 2 case - 213:32
Experiments15:07
Synthetic experiments16:31
Synthetic experiments: average regrets over 5000 experiments17:59
Real data experiments18:47
Real data experiment20:22
Discussion21:03
Conclusion and perspectives22:01