An acquisition of the relation between vision and action using self-organizing map and reinforcement learning.
Kazunori TeradaHideaki TakedaToyoaki NishidaPublished in: KES (1) (1998)
Keyphrases
- self organizing maps
- reinforcement learning
- action selection
- unsupervised learning
- competitive learning
- neural network
- computer vision
- action space
- topology preserving
- som neural network
- neural gas
- state space
- learning vector quantization
- function approximation
- k means
- input data
- feature maps
- vision system
- supervised learning
- growing self organizing map
- optimal policy
- markov decision processes
- learning algorithm
- learning process
- pattern classification
- clustering method
- database
- bayesian networks
- probabilistic neural network
- visualization methods
- agent learns