CMSC 691E: Emergence

Fall 2003, Tuesday/Thursday 11:30-12:45, FH 227
Prof. Marie desJardins

Class Description

This graduate seminar will explore the concept of emergence as it relates to AI and computational systems. Emergence is a broad concept that refers to systems in which simple local behaviors result in complex ("emergent") global behaviors. Emergence is studied in diverse fields, including multi-agent systems, biology, graphics, economics, physics, mathematics, and organization theory.

Topics will include computational and biological emergence, chaos theory, and dynamical network-based systems. Additional topics will be identified during the course of the semester; these may include agent-based simulation and swarm systems, particle systems, artificial life, cellular automata, and organizational dynamics.

The class will primarily be a discussion class, with assigned weekly readings and rotating discussion leaders. Each student will complete a hands-on design project using one of the available software packages for simulating emergent systems, write an in-depth paper on a topic within emergence, and give a presentation on this topic to the class. As a class, we will produce an annotated bibliography on the topic of emergence as it relates to AI and multi-agent systems, write a survey article on this topic (suitable for publication in AI Magazine or a journal such as Autonomous Agents and Multi-Agent Systems), and develop a set of well defined, open research problems that students might want to consider working on.



Reading list (evolving)


Thoughts from the class:

bullet WIKI
bullet Examples of emergent systems
bullet Properties of emergent systems