Online Reinforcement Learning Control by Direct Heuristic Dynamic Programming: from Time-Driven to Event-Driven.
Qingtao ZhaoJennie SiJian SunPublished in: CoRR (2020)
Keyphrases
- event driven
- dynamic programming
- reinforcement learning
- optimal control
- control problems
- state space
- real time
- optimal policy
- markov decision processes
- control system
- control strategies
- information delivery
- control policy
- robot control
- function approximation
- lagrangian relaxation
- publish subscribe
- dynamic programming algorithms
- learning algorithm
- policy search
- robotic control
- control strategy
- online learning
- genetic algorithm
- reinforcement learning algorithms
- action selection
- infinite horizon
- single machine
- greedy algorithm
- greedy search
- markov decision problems
- query processing
- search algorithm
- database systems