A two-stage reinforcement learning-based approach for multi-entity task allocation

Aicheng Gong, Kai Yang, Jiafei Lyu, Xiu Li

Abstract: Highlights•Innovative Task Allocation: Uses Markov Decision Process and reinforcement learning for better solutions.•Dynamic Adaptation: Attention mechanisms adapt to changing entities and tasks efficiently.•Fine-Tuned Selection: Seq2seq structure optimizes agent selection, reducing costs and boosting efficiency.