Value Function Approximation in Reinforcement Learning Using the Fourier Basis.
George Dimitri KonidarisSarah OsentoskiPhilip S. ThomasPublished in: AAAI (2011)
Keyphrases
- reinforcement learning
- state space
- temporal difference
- temporal difference learning
- function approximation
- approximate dynamic programming
- state action
- machine learning
- optimal policy
- frequency domain
- fourier transform
- policy search
- radon transform
- neural network
- reinforcement learning algorithms
- learning algorithm
- function approximators
- markov games
- fourier analysis
- multiscale
- policy iteration
- supervised learning
- action selection
- evaluation function
- basis functions