Detailed notes on Multi-agent Systems
By osinachi mbah
This notes describes the behaviour of multi-agents in an environment and their communial and competitve interactions among themselves. In reinforcement learning, an agent is placed in an environment to autonomously understand it. Rewards are given for an achieved objectives and penalties are awarded for failed goals. Imagine an environment with more than one agent. How do they interact between themselves and the environment to reduce the risk of getting penalties and obtain as much rewards as possible? This question is covered in these notes and more.
Outline:
Introduction to the concepts of autonomous agents and multiagent systems: intelligent autonomous agents.
Learning in determnistic environments: Dijkstra algorithms and A*-search algorithm.
Reinforcement learning, Markov decision processes and optimal policies: Value iteration, Q-learning and SARSA algorithm.
Interaction between agents: multiagent coordination.
Partially observable Markov Decision Processes: Hidden Markov Models
No reviews yet.