Classification on UCI letter (Iteration 20)
0.0029Metric GapAOE
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| AOEIterations=20, Candidate Models=10,000, Surrogate Model=GP binary classifier (Matérn 3/2), Epsilon-greedy step (ε)=0.05, Optimizer=Adam, Batch Size=100, Learning Rate=0.0012021.05 | 0.0029 | 0.011 | |
| DR-EIIterations=20, Candidate Models=10,000, Baseline Model=Logistic Regression, Epsilon-greedy step (ε)=0.052021.05 | 0.013 | 0.044 | |
| DR-gIterations=20, Candidate Models=10,000, Baseline Model=Logistic Regression, Epsilon-greedy step (ε)=0.052021.05 | 0.02 | 0.054 | |
| IS-gIterations=20, Candidate Models=10,000, Epsilon-greedy step (ε)=0.052021.05 | 0.024 | 0.063 | |
| IS-EIIterations=20, Candidate Models=10,000, Epsilon-greedy step (ε)=0.052021.05 | 0.042 | 0.061 | |
| BOIterations=20, Candidate Models=10,000, Epsilon-greedy step (ε)=0.052021.05 | 0.059 | 1.31 |