Community Research and Development Information Service - CORDIS

Abstract

Optimization of performance in collective systems often requires altruism. Emergence and stabilization of altruistic behaviours are difficult because the agents incur a cost when behaving altruistically. In this paper we propose a biologically inspired strategy to learn stable altruistic behaviours in artificial multi-agent systems, namely reciprocal altruism. Our multi-agent system is made up of autonomous agents with a behaviour-based architecture. Agents learn the most suitable cooperative strategy for different environments by means of a reinforcement learning algorithm. Each agent receives a reinforcement signal that only measures its individual performance. Simulation results show how the multi-agent systems learn stable altruistic behaviours, so reaching optimal (or near optimal) performances in unknown and changing environments.

Additional information

Authors: ZAMORA J, Universidad Complutense de Madrid, Facultad de Biología, Dpt. de Biomatemática (ES);MURCIANO A, Universidad Complutense de Madrid, Facultad de Biología, Dpt. de Biomatemática (ES);DEL R MILLÁN J, Universidad Complutense de Madrid, Facultad de Biología, Dpt. de Biomatemática (ES);DEL R MILLAN J, Universidad Complutense de Madrid, Facultad de Biologia, Dpt. de Biomatematica (ES)
Bibliographic Reference: Paper presented: IEEE International Symposium on Computational Intelligence in Robotics and Automation, Monterey (US) July 10-11, 1997
Availability: Available from (1) as Paper EN 40696 ORA
Record Number: 199711236 / Last updated on: 1997-10-10
Category: PUBLICATION
Original language: en
Available languages: en