Deadline-Aware Task Offloading With Partially-Observable Deep Reinforcement Learning for Multi-Access Edge Computing.
Hui HuangQiang YeYitong ZhouPublished in: IEEE Trans. Netw. Sci. Eng. (2022)
Keyphrases
- partially observable
- reinforcement learning
- markov decision processes
- state space
- partially observable domains
- dynamical systems
- partial observability
- partially observable environments
- decision problems
- partial observations
- hidden state
- function approximation
- markov decision problems
- reward function
- infinite horizon
- action models
- belief state
- optimal policy
- fully observable
- model free
- learning algorithm
- partially observable markov decision processes
- machine learning
- belief space
- optimal control
- temporal difference
- policy iteration
- finite state
- linear programming
- dynamic programming
- special case
- knowledge base