Towards Collective Learning for MAS
نویسندگان
چکیده
Yutaka I. Leon Suematsu, Keiki Takadama, Katsunori Shimohara, Osamu Katai 1 ATR Cognitive Information Science Laboratories, Kyoto, Japan 2 NiCT Universal Media Research Center, Kyoto, Japan 3 Graduate School of Informatics, Kyoto University, Kyoto, Japan 4 Tokyo Institute of Technology, Tokyo, Japan 5 Doshisha University, Kyoto, Japan [email protected], [email protected], [email protected], [email protected]
منابع مشابه
Evolutionary Multi-agent Systems
In Multi-Agent learning, agents must learn to select actions that maximize their utility given the action choices of the other agents. Cooperative Coevolution offers a way to evolve multiple elements that together form a whole, by using a separate population for each element. We apply this setup to the problem of multi-agent learning, arriving at an evolutionary multi-agent system (EA-MAS). We ...
متن کاملCOllective INtelligence with Sequences of Actions - Coordinating Actions in Multi-agent Systems
The design of a Multi-Agent System (MAS) to perform well on a collective task is non-trivial. Straightforward application of learning in a MAS can lead to sub optimal solutions as agents compete or interfere. The COllective INtelligence (COIN) framework of Wolpert et al. proposes an engineering solution for MASs where agents learn to focus on actions which support a common task. As a case study...
متن کاملA Framework for Learning Agents and its Application to Market-Based MAS
This document describes work in progress that aims towards the implementation of a framework for describing and understanding the behavior of systems with learning agents (that learn models of other agents), and the application and use of this framework within the market-based MAS that is the University of Michigan Digital Library.
متن کاملLEAF: A Toolkit for Developing Coordinated Learning Based MAS
This paper describes LEAF, the “Learning Agent based FIPA-Compliant Community Toolkit”, a toolkit for developing multiagent systems coordinated using utility function assignment, based on Collective Intelligence by Wolpert et al. LEAF agents use machine learning techniques such as reinforcement learning to maximise local utility functions, where local utility functions are assigned to agents su...
متن کاملDesigning Adaptive Systems Using Teleo-Reactive Agents
Although adaptivity is a central feature of agents and multiagent systems (MAS), there is no precise definition of it in the literature. What does it mean for an agent or for a MAS to be adaptive? How can we reason about and measure the ability of agents and MAS to adapt? How can we systematically design adaptive systems? In this paper, we provide a formal definition of adaptivity, and a framew...
متن کامل