Reinforcement Learning Integrated with a Non-Markovian Controller.
Istvan SzitaBálint TakácsAndrás LörinczPublished in: ECAI (2002)
Keyphrases
- reinforcement learning
- optimal control
- adaptive control
- reinforcement learning agents
- reinforcement learning algorithms
- actor critic
- state space
- function approximation
- control policy
- learning algorithm
- temporal difference
- control strategies
- machine learning
- control system
- model free
- markov decision processes
- dynamic programming
- learning classifier systems
- neural network
- reward function
- fuzzy logic
- control scheme
- control strategy
- optimal policy
- real time
- state abstraction
- partially observable domains
- decision processes
- control theory
- stochastic process
- initial state
- pid controller
- closed loop
- learning process