Learning in multi-agent systems

Abstract

In recent years, multi-agent systems (MASs) have received increasing attention in the artificial intelligence community. Research in multi-agent systems involves the investigation of autonomous, rational and flexible behaviour of entities such as software programs or robots, and their interaction and coordination in such diverse areas as robotics (Kitano et al., 1997), information retrieval and management (Klusch, 1999), and simulation (Gilbert & Conte, 1995). When designing agent systems, it is impossible to foresee all the potential situations an agent may encounter and specify an agent behaviour optimally in advance. Agents therefore have to learn from, and adapt to, their environment, especially in a multi-agent setting.

Emergent and Adaptive Systems of Systems

“…There exist a plethora of proposals in the literature on how to extend RL to MAS, all of which follow in one way or another the Free Market model described in Section II (see [11] for a survey). The simplest way to extend the single-agent Q-learning algorithm to multi-agent games is to add a subscript to the original formulae, that is, to have the learning agent pretend that the environment is passive.…”

SIMBA: A simulator for business education and research

“…Some business simulator games use a mixture of human and machine learning agents [20]. The learning agents can use a typical genetic based learning classifier system, XCS (eXtended learning Classifier System) [1]. In that study, the authors developed four kinds of agents as alternatives to human players.…”

“…Reinforcement learning is widely used in multi agent systems in order to improve the behavior of the agents [1]. Among many different reinforcement learning algorithms, Q learning [36,37] has been widely used.…”

Design of Multi-Behavior Agents for Supply Chain Planning: An Application to the Lumber Industry

“…Agents then have the possibility of recognizing situations and applying the best behavior instead of trying each of them one at the time. Alonso (Alonso et al, 2001) argues that learning is the most crucial characteristic of intelligent agent systems. Many researchers have been investigating learning agents, from defining fundamental issues of intelligent learning agents (Schleiffer, 2005) to describing major learning techniques for multi-agents systems (Alonso et al, 2001;Weiss & Sen, 1996).…”

“…Alonso (Alonso et al, 2001) argues that learning is the most crucial characteristic of intelligent agent systems. Many researchers have been investigating learning agents, from defining fundamental issues of intelligent learning agents (Schleiffer, 2005) to describing major learning techniques for multi-agents systems (Alonso et al, 2001;Weiss & Sen, 1996). Shen (Shen et al, 2000) present a research review related to the enhancement of agent-based manufacturing systems through learning, including the use of learning in a more general manufacturing context.…”