Goal-Directed Reinforcement Learning Using Variable Learning Rate.
Arthur Plínio de S. BragaAluizio F. R. AraújoPublished in: SBIA (1998)
Keyphrases
- goal directed
- learning rate
- reinforcement learning
- learning algorithm
- autonomous learning
- error function
- convergence rate
- hidden layer
- convergence speed
- rapid convergence
- autonomous robots
- multilayer neural networks
- adaptive learning rate
- markov decision processes
- weight vector
- function approximation
- reinforcement learning algorithms
- state space
- convergence theorem
- machine learning
- real time
- optimal control
- model free
- temporal difference
- high accuracy
- reinforcement learning methods
- search space
- genetic algorithm