RMDDQN-Learning: Computation Offloading Algorithm Based on Dynamic Adaptive Multi -Objective Reinforcement Learning in Internet of Vehicles.
Xiangjun ZhangWeiguo WuZhihe ZhaoJinyu WangSong LiuPublished in: IEEE Trans. Veh. Technol. (2023)
Keyphrases
- learning algorithm
- reinforcement learning
- multi objective
- optimization algorithm
- dynamic programming
- detection algorithm
- actor critic
- computational complexity
- search space
- cost function
- model based reinforcement learning
- learning process
- supervised learning
- particle swarm optimization
- model free
- np hard
- dynamic environments
- path planning
- learning tasks
- convergence rate
- active learning
- learning capabilities
- linear predictors
- policy gradient
- optimal solution
- objective function
- multiagent evolutionary algorithm