DL-DRL: A double-layer deep reinforcement learning approach for large-scale task scheduling of multi-UAV.
Xiao MaoGuohua WuMingfeng FanPublished in: CoRR (2022)
Keyphrases
- reinforcement learning
- description logics
- multiple layers
- learning algorithm
- deep learning
- path planning
- real world
- application layer
- resource management
- expressive power
- small scale
- function approximation
- knowledge base
- computing systems
- multi layer
- vision based navigation
- transfer learning
- supervised learning
- markov decision processes
- optimal policy
- model free
- state space
- learning process
- real life
- policy search
- machine learning