Q-learning with exploration driven by internal dynamics in chaotic neural network.
Toshitaka MatsukiSouya InoueKatsunari ShibataPublished in: IJCNN (2020)
Keyphrases
- chaotic neural network
- chaotic dynamics
- phase space
- action selection
- cellular automata
- activation function
- hopfield neural network
- reinforcement learning
- dynamical systems
- sufficient conditions
- state space
- learning rate
- learning algorithm
- dynamic model
- neural network
- dynamic programming
- artificial neural networks
- image segmentation