Table 9 Experimental parameters of improved Q-Learning algorithm
From: Optimal scheduling in cloud healthcare system using Q-learning algorithm
parameters | values |
|---|---|
Learning rate α | 0.1 |
discounted factor γ | 0.9 |
iteration times T | 500 |
ε | \(\upvarepsilon =0.5/\left(1+{e}^{\frac{10\times \left(\text{ episode} - 0.6\times \mathrm{max}\_{\text{episode}}\right)}{\text{ max\_episodes}}}\right)\) |