Neural Network Approximation for Pessimistic Offline Reinforcement Learning.
Di WuYuling JiaoLi ShenHaizhao YangXiliang LuPublished in: AAAI (2024)
Keyphrases
- neural network
- reinforcement learning
- back propagation
- function approximation
- approximation algorithms
- artificial neural networks
- function approximators
- state space
- neural nets
- neural network model
- approximation error
- fuzzy logic
- multi agent
- pattern recognition
- learning capabilities
- multilayer perceptron
- learning algorithm
- real time
- network architecture
- radial basis function
- approximation methods
- network model
- model free
- temporal difference
- reinforcement learning algorithms
- markov decision process
- error bounds
- transfer learning
- decision making
- recurrent neural networks
- feed forward
- multi layer
- closed form
- activation function
- supervised learning
- dynamic programming
- special case
- feed forward neural networks
- control system
- objective function
- genetic algorithm