Binary Classification on Higgs
0.2794Lossreference architecture
Evaluation results
| Method | Links | |
|---|---|---|
| reference architectureArchitecture=16-144-16, Number of Parameters=5,2652022.04 | 0.2794 | |
| RL with rejection-based rewardArchitecture=64-16-128, Number of Parameters=5,2652022.04 | 0.287 | |
| RL with Abs RewardArchitecture=88-16-64, Number of Parameters=5,2172022.04 | 0.2874 | |
| RENAArchitecture=80-8-152, Number of Parameters=4,5692022.04 | 0.2882 | |
| RL with Abs RewardArchitecture=80-16-88, Number of Parameters=5,2812022.04 | 0.2887 | |
| MNasNet (reward type 1)Architecture=64-32-16, Number of Parameters=4,5452022.04 | 0.2888 | |
| RL with rejection-based rewardArchitecture=80-32-8, Number of Parameters=5,2652022.04 | 0.2893 | |
| RL with Abs RewardArchitecture=104-16-24, Number of Parameters=5,2332022.04 | 0.2896 | |
| MNasNet (reward type 1)Architecture=72-24-32, Number of Parameters=4,7452022.04 | 0.2898 | |
| MNasNet (reward type 1)Architecture=72-24-24, Number of Parameters=4,5372022.04 | 0.2903 | |
| RL with rejection-based rewardArchitecture=72-24-48, Number of Parameters=5,1612022.04 | 0.2911 | |
| MNasNet (reward type 2)Architecture=16-8-16, Number of Parameters=7772022.04 | 0.2929 | |
| MNasNet (reward type 2)Architecture=8-8-24, Number of Parameters=5532022.04 | 0.3003 |