Measuring the Impact of Memory Replay in Training Pacman Agents using Reinforcement Learning.
Fabian Fallas-MoyaJeremiah DuncanTabitha K. SamuelAmir SadovnikPublished in: CLEI (2021)
Keyphrases
- reinforcement learning
- multi agent
- learning agents
- multi agent systems
- intelligent agents
- multi agent reinforcement learning
- action selection
- multiagent systems
- single agent
- multiple agents
- agent receives
- learning agent
- software agents
- autonomous agents
- supervised learning
- cooperative
- multiagent learning
- function approximation
- optimal policy
- multi agent environments
- reinforcement learning agents
- evolutionary learning
- artificial agents
- multiagent reinforcement learning
- agent systems
- decision making
- learning capabilities
- reinforcement learning algorithms
- memory usage
- memory requirements
- state space
- model free
- agent architecture
- incomplete information
- resource allocation
- dynamic environments
- training set
- learning algorithm
- past experience
- agent technology
- partial observability
- multi agent learning
- transfer learning