Computer Science (COMP) 667

Overview

Multiagent systems (MAS) can be defined as loosely coupled networks of problem solvers that interact to solve problems that are beyond the individual capabilities or knowledge of each problem solver. These problem solvers, often called agents, are autonomous and can be heterogeneous in nature.

Research and development in MAS is concerned with the study, behavior, and construction of a collection of possibly pre-existing autonomous agents that interact with each other and their environments. The study of such systems goes beyond the study of individual intelligence in its consideration of problem solving with social components.

COMP 667 introduces students to the main topics in the theory and practice of intelligent agents and MAS – currently one of the most important and rapidly expanding areas of computer science, having emerged from the study of distributed artificial intelligence. Multiagent systems have been used as an important means with which to address the development of large and complex information systems (IS), Computer Supported Cooperative Work (CSCW) environments, and Decision Support Systems (DSS).

Agent programming is also a key component of agent-oriented software engineering (AOSE), which enhances the ability of traditional software development practice. This course will introduce the student to the concept and architecture of agents. This course will also explain agent-oriented analysis, design, and programming (AOP), and the principles and methodologies supporting the realization of agent-based systems.

Course Objectives

• Distinguish between the major approaches to building agents.
• Describe the protocols for agent coordination.
• Develop strategies for using protocols and reasoning capabilities to realize the benefits of cooperation.
• Describe the concepts and algorithms that comprise the foundations of distributed problem solving and planning.
• Compare various search algorithms for agents.
• Analyze methods for making socially desirable decisions among rational agents.
• Analyze the typical approaches to learning in multiagent systems (MAS).
• Review, judge, and infer potential directions of research in learning in MAS.
• Explain the formal bases for abstractions and constructions that arise in the study of agents and MAS, and explain how these concepts may be realized in a practical interpreter.
• Analyze successful industrial and practical applications of MAS.
• Apply and compare various MAS analysis and design methodologies, and computer-aided development tools.

Learning Outcomes

While completing this course, students will

• learn the concepts and theories of MAS to gain new knowledge in this field;
• develop the competency and skills to construct MAS, both individually as well as in teams;
• develop innovative strategies, methodologies, and techniques to address issues in analysis, design, and implementation of MAS;
• compare and contrast alternative methodologies, models, and tools used in the design and development of MAS;
• develop business and real-world perspectives of MAS;
• be aware of future and current trends in MAS research and applications;
• effectively communicate course work in writing and oral presentation.

Outline

• Unit 1: Intelligent Agents
• Unit 2: Multiagent Systems
• Unit 3: Distributed Problem Solving and Planning
• Unit 4: Search Algorithms for Agents
• Unit 5: Group Decision-Making
• Unit 6: Multiagent System Design and Programming
• Unit 7: Learning in Multiagent Systems
• Unit 8: Logic-Based Representation and Reasoning
• Unit 9: Industrial Deployment of Multiagent Systems

Evaluation

In order to receive credit for COMP 667, you must achieve a cumulative course grade of "C+" (66 percent) or better, complete all three assignments, must achieve an average grade of at least 60 percent on the assignments and achieve a grade of at least 60 percent on the Project. Your cumulative course grade will be based on the following assessment.

The weighting of the composite grade is as follows: