Table 3 Hyperparameters of NPER-D3QN.
From: Latency and energy-aware adaptive service migration in mobile edge computing
Parameter | Symbol | Value |
|---|---|---|
Episode | M | 500 |
Time slots | T | 80 |
Replay Memory | N | 10000 |
Batch size | n | 64 |
Reward discount factor | \(\gamma\) | 0.99 |
Initial value of noise parameter | \(\sigma\) | 0.017 |
Target network update frequency | C | 50 |
Learning rate | \(\psi\) | 0.001 |