Graph convolutional recurrent networks for reward shaping in reinforcement learning.
Hani SamiJamal BentaharAzzam MouradHadi OtrokErnesto DamianiPublished in: Inf. Sci. (2022)
Keyphrases
- reward shaping
- reinforcement learning
- recurrent networks
- reinforcement learning algorithms
- complex domains
- recurrent neural networks
- feed forward
- markov decision problems
- neural network
- function approximation
- state space
- biologically inspired
- dynamic programming
- markov decision processes
- model free
- temporal difference
- random walk
- learning agent
- artificial neural networks
- transition model
- action selection
- reward function
- transfer learning
- machine learning