EE Department Intranet -
Close window CTRL+W

ELEC97105 (EE4-70) Self-Organising Multi-Agent Systems

Lecturer(s): Prof Jeremy Pitt


This course will introduce students to the theories, concepts and technologies in two fields of study at the intersection of Artificial Intelligence and Distributed Computing, namely Multi-Agent Systems and Self-Organising Systems.

The curriculum is intentionally broad to expose final year students to the considerable variety of theories, and to provide an inter-disciplinary overview, but with sufficient depth to appreciate the complexity and timeliness of the subject; and demonstrate that these topics are the subject of intensive research and development, both in the college and around the world, and are at the forefront of engineering many applications.

Learning Outcomes

Students will be able to understand, from a computational perspective, systems of autonomous components interacting in the context of conventional rules, i.e. self-organising multi-agent systems. This understanding will cut across three dimensions: firstly, foundational, with respect to basic concepts of computational agency and self-organisation; secondly, how to analyse strategic decision problems of action selection, preference selection and behaviour regulation; and thirdly, how to resolve social problems, such as achieving fairness, managing knowledge and maintaining order, using conceptual resources.

Students will be able to apply this knowledge to the principles and practices underpinning the design and implementation of data structures and algorithms for collective decision-making, distributed knowledge management, and coordinated action, in the context of:
* individual preferences and choices, ...
* ... made by autonomous entities, ...
* ... exhibiting some form of 'intelligence', ...
* ... directed towards prosocial or socially-productive purposes, ...
* ... in the context of institutional (i.e. rule-based) structures and processes.


Part 1: Foundations
Agents & Multi-Agent Systems: agents and agent architectures, multi-agent systems, agent communication and institutionalised power, agent societies and Self-Adaptation: autonomous systems, autonomic systems, dynamic norm-governed systems
Self-Organisation: swarms and cellular automata, dimensions of change, planned emergence, parameter reconfiguration and rule reconfiguration

Part 2: Strategic Interaction
Game Theory: 2-player sequential games, dominant strategy, Nash Equilibrium, mechanism design, Evolutionary GT
Social Choice Theory: voting rules and winner determination, manipulation, Arrow’s Theorem and other paradoxes, voting protocol
Alternative Disupute Resolution: error tolerance, error correction, dispute resolution
Ostrom Institution Theory: electronic institutions, Ostrom’s principles, collective action and sustainability, system of systems

Part 3: Social Interaction
Computational Justice: justice and fairness, algorithms for distributive, retributive and procedural justice
Social Construction: conceptual resources: trust, forgiveness and social capital; alternative economies, values, artificial social construction and computational axiology
Knowledge Aggregation: knowledge management, opinion formation, social network interactional justice
Algorithmic Self-Governance: basic democracy, democracy by design
Exam Duration: N/A
Coursework contribution: 100%

Term: Autumn

Closed or Open Book (end of year exam): N/A

Coursework Requirement:
         Coursework only module

Oral Exam Required (as final assessment): N/A

Prerequisite module(s): None required

Course Homepage:

Book List: