Return
Dynamic grouping evolutionary reinforcement learning algorithm for scheduling workflows in hybrid clouds
H
B
黄
B
J
Y
DOI:10.1016/j.asoc.2026.116185.png)
Abstract
En 中文
• A swarm-guided and gradient-refined ERL framework by combining DDPG and PSO. • A dynamic group learning-enabled update mechanism to reduce time complexity. • A dual-population enhanced and hybrid elite solution-guided evolution strategy. • A complementary heuristic-guided action selection to trade-off two objectives.
Journal
IF:
6.6
Papers:
1.4W
Citations:
4.8W
