Q-map: a Convolutional Approach for Goal-Oriented Reinforcement Learning.
Fabio PardoVitaly LevdikPetar KormushevPublished in: CoRR (2018)
Keyphrases
- goal oriented
- reinforcement learning
- process oriented
- function approximation
- decentralized control
- transfer learning
- markov decision processes
- maximum a posteriori
- deep learning
- optimal control
- state space
- optimal policy
- data marts
- requirements analysis
- model free
- reinforcement learning algorithms
- learning classifier systems
- neural network
- learning algorithm
- requirements engineering
- e learning
- temporal difference
- action selection
- multi agent
- sparse coding
- multiagent systems
- fine grained
- supervised learning