Learning to Communicate to Solve Riddles with Deep Distributed Recurrent Q-Networks.
Jakob N. FoersterYannis M. AssaelNando de FreitasShimon WhitesonPublished in: CoRR (2016)
Keyphrases
- learning process
- multi agent
- supervised learning
- connectionist networks
- neural network
- spiking neural networks
- computer networks
- learning tasks
- active learning
- prior knowledge
- reinforcement learning
- case study
- learning algorithm
- cooperative
- peer to peer
- knowledge acquisition
- unsupervised learning
- learning systems
- learning problems
- communication cost
- knowledge base