ScaRLib: Towards a hybrid toolchain for aggregate computing and many-agent reinforcement learning.
Davide DominiFilippo CavallariGianluca AguzziMirko ViroliPublished in: Sci. Comput. Program. (2024)
Keyphrases
- reinforcement learning
- multi agent
- action selection
- multi agent systems
- intelligent agents
- autonomous agents
- learning agent
- multiagent systems
- reinforcement learning algorithms
- reward function
- decision making
- learning capabilities
- multiple agents
- function approximation
- software agents
- single agent
- reward shaping
- mobile agents
- agent architecture
- reinforcement learning agents
- exploration strategy
- autonomous learning
- learning agents
- state action
- markov decision process
- markov decision processes
- partially observable
- dynamic programming
- optimal policy
- agent model
- state space
- dynamic environments
- multiagent reinforcement learning
- reasoning process
- agent systems
- machine learning
- action space
- supervised learning
- temporal difference
- multi agent reinforcement learning
- state abstraction
- optimal control