Computer Science (COMP) 657

Overview

Artificial Intelligence (AI) aims to make computers and information systems more "intelligent" to solve complex problems and provide more natural and effective services to human beings. AI has been a source of innovative ideas and techniques in computer science, and has been widely applied to many information systems. This course provides a comprehensive, graduate-level introduction to artificial intelligence, emphasizing advanced topics such as advanced search, reasoning and decision-making under uncertainty, and machine learning.

The best way to study AI in this course is to learn by researching. In addition to a comprehensive introduction to a variety of AI subfields, this course will facilitate your exploration of state-of-the-art research and applications of AI, the production of your own ideas and visions, and discussions and sharing of some deep understandings. You will be asked to think about and communicate higher-level abstract ideas, including the concepts, strategies, principles, and algorithms of AI, rather than the technical details of implementation and programming. You will also be encouraged to test and evaluate ideas and algorithms through research-oriented studies involving software programming and/or exploring.

The course is divided into four parts. Part I: Advanced Searching, concentrates on informed search, online search and adversarial search. Part I also includes constraint satisfaction problems (CSP), and optimal adversarial searching and exploration.

Part II: Knowledge, Reasoning, and Decision Making under Uncertainty, discusses how to represent knowledge,including incomplete and uncertain knowledge of the real world; how to reason logically with that knowledge using probabilities; how to use these reasoning models and methods to decide what to do, particularly by constructing plans; and how to reason and make decisions in the presence of uncertainty about the world. It includes some state-of-the-art topics, such as the logical representation of different types of knowledge, reasoning under uncertainty, ontological engineering, hidden Markov models (HMM), dynamic Bayesian networks (DBN), and graphical models.

Part III: Machine Learning, describes both symbolic and statistical learning methods, as well as reinforcement learning, deep learning and multi-agent learning for generating the knowledge required by reasoning and/or decision-making components of intelligent agents or systems.

Part IV: AI for Communication, Robotics, and Information Systems, introduces some of the most important aspects related to human activities, including natural language, perception, robotics, and human-centered information systems such as business intelligence and e-learning.

Course Objectives

The objectives of this course are to provide graduate students of M.Sc. Information Systems with comprehensive and in-depth knowledge of AI principles and techniques by introducing AI’s fundamental problems, and the state-of-the-art models and algorithms used to undertake these problems. This course is also designed to expose students to the frontiers of AI-intensive computing and information systems, while providing a sufficiently strong foundation to encourage further research.

Learning Outcomes

After completing this course, students will be able to:

• Compare AI with human intelligence and traditional information processing, and discuss its strengths and limitations and its application to complex and human-centered problems.
• Discuss the core concepts and algorithms of advanced AI, including informed searching, CSP, logic, uncertain knowledge and reasoning, dynamic Bayesian networks, graphical models, decision making, multiagent, inductive learning, statistical learning, reinforcement learning, deep learning, natural language processing, robotics, and so on.
• Apply the basic principles, models, and algorithms of AI to recognize, model, and solve problems in the analysis and design of information systems.
• Analyze the structures and algorithms of a selection of techniques related to searching, reasoning, machine learning, and language processing.
• Design AI functions and components involved in intelligent systems, such as computer games, expert systems, semantic web, information retrieval, machine translation, mobile robots, decision support systems, and intelligent tutoring systems.
• Review research articles from well-known AI journals and conference proceedings regarding the theories and applications of AI.
• Perform research project and write research proposal, report and paper.

Outline

Part I: Advanced Searching

1. Informed Search and Online Search
2. Constraint Satisfaction Problems (CSP) and Distributed CSP
3. Games and Adversarial Search

Part II: Knowledge, Reasoning, and Decision Making under Uncertainty

1. Logic, Inference, and Ontology
2. Automated Planning and Acting (Optional)
3. Uncertainty, Graphical Models, and Probabilistic Reasoning
4. Temporal Probabilistic Reasoning and Dynamic Bayesian Networks
5. Complex Decision-Making

Part III: Machine Learning

1. Inductive and Analytic Learning
2. Computational Learning Theory and Statistical Learning
3. Reinforcement Learning, Deep Learning and Multiagent Learning

Part IV: AI for Communication, Robotics, and Information Systems

1. Natural Language Understanding and Statistical Language Processing
2. Perception and Robotics (Optional)
3. AI in Business Intelligence and E-Learning (Optional)

The units or sections marked as Optional in the Syllabus and Study Guide of this course contain content that is not required for the Final Exam. However, the students who are interested in or familiar with these topics are encouraged to explore them, and to use the material and applications to complete their TMAs and final examination.

Evaluation

To receive credits for COMP 657 toward the Master of Science in Information Systems Program, you must achieve a cumulative course grade of at least B- (70 percent), including an average grade of 60 percent on the assignments and achieve a grade of at least 60 percent on the Final Examination.

To receive credit for COMP 657 toward the Post-Baccalaureate Certificate in Data Analytics Program, you must achieve a cumulative course grade of at least C+ (66 percent), including an average grade of 60 percent on the assignments and achieve a grade of at least 60 percent on the Final Examination.

To receive credit for COMP 657 as a non-program student, you must achieve a cumulative course grade of at least C+ (66 percent), including an average grade of 60 percent for each required learning outcome it is intended to address.