Reinforcement Learning on CartPole (Wall-clock Training Time)
0.1Wall-clock Training Time (min)Independent QHD
Evaluation results
| Method | Links | |
|---|---|---|
| Independent QHDClient heterogeneity setup=Heterogeneous, Number of episodes=600, Number of clients=52026.05 | 0.1 | |
| Truncate FedAvg-QHDClient heterogeneity setup=Heterogeneous, Number of episodes=600, Number of clients=52026.05 | 0.3 | |
| FedQHDClient heterogeneity setup=Heterogeneous, Number of episodes=600, Number of clients=52026.05 | 1.6 | |
| Independent QHDClient heterogeneity setup=Homogeneous, Number of episodes=600, Number of clients=52026.05 | 1.7 | |
| Truncate FedAvg-QHDClient heterogeneity setup=Homogeneous, Number of episodes=600, Number of clients=52026.05 | 3.5 | |
| Oracle QHD†Client heterogeneity setup=Homogeneous, Number of episodes=600, Number of clients=52026.05 | 6.1 | |
| Oracle QHD†Client heterogeneity setup=Heterogeneous, Number of episodes=600, Number of clients=52026.05 | 6.1 | |
| FedQHDClient heterogeneity setup=Homogeneous, Number of episodes=600, Number of clients=52026.05 | 7.8 | |
| Oracle DQN†Client heterogeneity setup=Homogeneous, Number of episodes=600, Number of clients=52026.05 | 8.9 | |
| Distillation FedDQNClient heterogeneity setup=Heterogeneous, Number of episodes=600, Number of clients=52026.05 | 10.3 | |
| Distillation FedDQNClient heterogeneity setup=Homogeneous, Number of episodes=600, Number of clients=52026.05 | 12.4 | |
| FedAvg-DQNClient heterogeneity setup=Homogeneous, Number of episodes=600, Number of clients=52026.05 | 25.9 | |
| Oracle DQN†Client heterogeneity setup=Heterogeneous, Number of episodes=600, Number of clients=52026.05 | 64.1 |