Adaptive Agents and Multi-Agent Systems III. Adaptation and Multi-Agent Learning

Traditionally multi-agent learning is considered as the intersection of two subfields of artificial intelligence: multi-agent systems and machine learning. Conventional machine learning involves a single agent that is trying to maximise some utility function without any awareness of existence of other agents in the environment (Mitchell, 1997). Meanwhile, multi-agent systems consider mechanisms for the interaction of autonomous agents. Learning system is defined as a system where an agent learns to interact with other agents (e.g., Clouse, 1996; Crites & Barto, 1998; Parsons, Wooldridge & Amgoud, 2003). There are two problems that agents need to overcome in order to interact with each other to reach their individual or shared goals: since agents can be available/unavailable (i.e., they might appear and/or disappear at any time), they must be able to find each other, and they must be able to interact (Jennings, Sycara & Wooldridge, 1998).

Download Full-text

FuzzyOrganization of Self-Adaptive Agents Based On Software Components

International Journal of Intelligent Information Technologies ◽

10.4018/ijiit.2014070103 ◽

2014 ◽

Vol 10 (3) ◽

pp. 36-56 ◽

Cited By ~ 2

Author(s):

Abderrahim Siam ◽

Ramdane Maamri ◽

Zaïdi Sahnoun

Keyword(s):

Membership Function ◽

Software Components ◽

Automatic Assembly ◽

Multi Agent Systems ◽

Adaptive Agents ◽

Agent Systems ◽

Proposed Model ◽

Multi Agent ◽

The Membership Function ◽

Self Adaptive

This paper addresses the development of organizational multi agent systems as a preferred solution to develop open, distributed and adaptive application. It proposes a combination between components and agents to define a flexible organizational model of MAS based on three concepts: roles, self-adaptive agents based on components and fuzzy groups. Roles are played by agents in fuzzy groups. A fuzzy group is a fuzzy set of agents characterized by a membership function expressing the partial membership of each agent to the group. The membership function expresses the degree of capacity of each agent to play a role. This work proposes a fuzzy measure of the capacity of agents to play roles. It proposes a model of auto adaptive agents constructed by automatic assembly (reassembly) of software components. Components implement required capabilities to play roles. The proposed model and introduced concepts have been tested using the Madkit platform.

Download Full-text

Learning Systems Engineering

Online and Distance Learning ◽

10.4018/978-1-59904-935-9.ch114 ◽

2011 ◽

pp. 1429-1438

Author(s):

Valentina Plekhanova

Keyword(s):

Machine Learning ◽

Systems Engineering ◽

Autonomous Agents ◽

Single Agent ◽

Learning System ◽

Multi Agent Systems ◽

Agent Systems ◽

Agent Learning ◽

Multi Agent ◽

Conventional Machine

Traditionally multi-agent learning is considered as the intersection of two subfields of artificial intelligence: multi-agent systems and machine learning. Conventional machine learning involves a single agent that is trying to maximise some utility function without any awareness of existence of other agents in the environment (Mitchell, 1997). Meanwhile, multi-agent systems consider mechanisms for the interaction of autonomous agents. Learning system is defined as a system where an agent learns to interact with other agents (e.g., Clouse, 1996; Crites & Barto, 1998; Parsons, Wooldridge & Amgoud, 2003). There are two problems that agents need to overcome in order to interact with each other to reach their individual or shared goals: since agents can be available/unavailable (i.e., they might appear and/or disappear at any time), they must be able to find each other, and they must be able to interact (Jennings, Sycara & Wooldridge, 1998).

Download Full-text