Graph Neural Networks and Reinforcement Learning for Behavior Generation in Semantic Environments.
Patrick HartAlois C. KnollPublished in: IV (2020)
Keyphrases
- neural network
- reinforcement learning
- multi agent environments
- graph theory
- learning capabilities
- function approximators
- artificial neural networks
- function approximation
- robot behavior
- pattern recognition
- fuzzy logic
- random walk
- real world
- graph representation
- autonomous robots
- graph model
- domain specific
- state space
- multi agent
- markov decision processes
- directed graph
- high level
- training process
- dynamic programming
- action selection
- feed forward
- learning algorithm
- connected components
- semantic information
- genetic algorithm
- dynamic environments
- machine learning
- heterogeneous environments
- learning classifier systems
- natural language
- co occurrence
- fuzzy systems
- weighted graph
- bipartite graph
- semantic network
- graph matching
- semantic similarity
- neural network model
- semantic annotation
- structured data
- back propagation
- semantic web