Deep Decentralized Multi-task Multi-Agent Reinforcement Learning under Partial Observability.
Shayegan OmidshafieiJason PazisChristopher AmatoJonathan P. HowJohn VianPublished in: CoRR (2017)
Keyphrases
- multi agent reinforcement learning
- multi task
- partial observability
- learning agent
- learning tasks
- reinforcement learning
- multi agent
- transfer learning
- learning problems
- cooperative
- partially observable
- machine learning
- learning algorithm
- state space
- single agent
- planning problems
- multi agent systems
- machine learning algorithms
- partially observable markov decision processes
- solving problems
- learning capabilities
- learning experience
- neural network
- belief state
- optimal policy
- feature selection
- reinforcement learning algorithms
- multiple agents
- learning process
- active learning
- kernel methods
- dynamic environments