Fall 2011 CS Course Listings

This file contains the Fall 2011 course listings for the Department of Computer Science and Engineering.

Archive of past courses

Course code: COMP5212
Course title: Machine Learning
Instructor: James Kwok
Room: 3519
Telephone: 2358-7013
Email:
WWW page:

Area in which course can be counted: AI

Course description: Machine learning is the science of making computer artifacts improve their performance without requiring humans to program their behavior explicitly. Machine learning has accomplished successes in a wide variety of challenging applications, ranging from computational molecular biology to computer vision to social web analysis. This course is a postgraduate-level introductory course in machine learning with emphasis put on the computational and mathematical principles underlying the most common machine learning problems and methods. It is not only suitable for students pursuing or planning to pursue research in machine learning or other related areas that focus on model and algorithm development, but is also suitable for students who want to apply disciplined machine learning techniques competently to their application-oriented research areas.

Background needed:
Computer science: object-oriented programming and data structures, design and analysis of algorithms
Mathematics: multivariable calculus, linear algebra and matrix analysis, probability and statistics

Course objective:
By the end of this course, students are expected to demonstrate ample competence in the following:
- Ability to take a real-world application and formulate the learning problems involved in it by identifying the major learning-related issues
- Ability to choose and apply the most common methods available for each of the major learning problem types
- Ability to compare different machine learning methods according to common performance criteria
- Ability to design and conduct empirical studies in such a way that the experimental results can be interpreted in accordance with disciplined scientific and statistical principles
- Ability to understand the motivations behind and the key issues studied in some recent research topics in machine learning

Course outline/content (by major topics):
** TENTATIVE TOPICS **
Supervised learning
Bayesian decision theory
Parameter estimation
Multivariate methods
Dimensionality reduction
Clustering
Nonparametric methods
Decision trees
Linear discrimination
Multilayer perceptrons
Support vector machines and kernel methods
Bayesian methods
Performance evaluation and comparison
Ensemble learning
Reinforcement learning
Quick overview of selected recent topics

Textbooks: None

Reference books/materials:
1. Christopher M. Bishop (2006). Pattern Recognition and Machine Learning. Springer.
2. Richard O. Duda, Peter E. Hart, and David G. Stork (2001). Pattern Classification. Second Edition. Wiley.
3. Trevor Hastie, Robert Tibshirani, and Jerome Friedman (2009). The Elements of Statistical Learning. Second Edition. Springer.
4. Tom M. Mitchell (1997). Machine Learning. McGraw-Hill.
5. Ethem Alpaydin (2010). Introduction to Machine Learning. Second Edition. MIT Press.
6. Other assigned reading material.

Grading scheme:
Based on class participation, homeworks and exams

Available for final year UG students to enroll: Yes

Minimum CGA required for UG students: Permission of the instructor

Course code: COMP5213 (Cancellation)
Course title: Introduction to Bayesian Networks
Instructor: Nevin L. Zhang
Room: 3504
Telephone: 2358-7015
Email:
WWW page: http://cse.hkust.edu.hk/~lzhang/

Area in which course can be counted: AI

Course description:
Bayesian networks (BNs) are a framework for dealing with the complexity that arises when applying probability theory in complex systems. They represent the structure of a system using a directed graph of random variables. Conditional independencies can be readily identified from the graph and are used to drastically reduce the complexity of inference. Model construction can be done manually using the intuitively appealing graphical interface provided. Alternatively, models can be learned from data via statistical principles such as maximum likelihood estimation and Bayesian estimation. The latter is the focus of much recent research and has attracted much attention in the AI, machine learning, and data mining communities.

The course consists of lectures by the instructor (75%) and student presentations (25%). The lectures are designed to provide a solid training in the theory and methods of BNs, while the student presentations are designed to convey an overview of the field.

Course objective:
To provide a solid training in Bayesian networks and related models for machine learning.

Course outline/content (by major topics):
     * Preparation:
           o Introduction to course
           o Multivariate Probability and Information Theory
     * Basic concepts:
           o Bayesian networks
           o D-separation
     * Inference:
           o Variable elimination
           o Clique tree propagation
     * Parameter learning:
           o Complete data
           o Incomplete data
     * Structure learning:
           o Structure Learning
           o Learning Latent structures
     * Applications:
           o BN for density estimation, classification, and clustering
           o Applications in Medicine
     * Term project presentations
           o Related models for machine learning

Textbooks: None

Reference books/materials:
    * D. Koller and N. Friedman (2009). Probabilistic Graphical Models: Principles and Techniques. edited by MIT Press.

Grading scheme:
    * Class participation: 10
    * Final exam: 40
    * Project report: 40
    * Oral presentation: 10

Available for final year UG students to enroll: Yes

Minimum CGA required for UG students: A-

Course code: COMP5311
Course title: Database Architecture and Implementation
Instructor: Wilfred Ng
Room: 3505
Telephone: 2358-6979
Email:
WWW page:

Area in which course can be counted: DB

Course description:
This course introduces basic concepts and implementation techniques in database management systems: disk and memory management; advanced access methods; implementation of relational operators; query processing and optimization; concurrency control and recovery.

Course objective:
Introductory database class for graduate students. The students are expected to learn basic concepts and implementation techniques of relational databases and advanced RDBMS applications.

Course outline/content(by major topics):
The instructor will teach the majority of the classes. Students will form groups. Each group will choose a general database area and prepare a presentation.

Textbooks:
Textbook: Database System Concepts, 5th Edition. A. Silberschatz, H. Korth, and S. Sudarshan.

Reference books/materials:
Reference: Database Management Systems, 3rd Edition. Raghu Ramakrishnan and Johannes Gehrke.

Grading scheme:
Student Presentations 20%, Midterm 35%, Final 45%. The exams will be with open books (any book or notes) and will be based on material explicitly covered during the classes.

Available for final year UG students to enroll: No

Course code: COMP5331
Course title: Knowledge Discovery in Databases
Instructor: Raymond Chi-Wing Wong
Room: 3542
Telephone: 2358-6982
Email:
WWW page: http://cse.hkust.edu.hk/~raywong/

Area in which course can be counted: DB or AI

Course description:
Data mining has emerged as a major frontier field of study in recent years. Aimed at extracting useful and interesting patterns and knowledge from large data repositories such as databases and the Web, the field of data mining integrates techniques from database, statistics and artificial intelligence. This course will provide a broad overview of the field, preparing the students with the ability to conduct research in the field.

Course objective:
To learn the techniques used in data mining research. To help the students get ready for research.

Course outline/content (by major topics):
1. Association
2. Clustering
3. Classification
4. Data Warehouse
5. Data Mining over Data Streams
6. Web Databases
7. Multi-criteria Decision Making

Textbooks:
Data Mining: Concepts and Techniques. Jiawei Han and Micheline Kamber. Morgan Kaufmann Publishers (2nd edition).

Reference books/materials:
Introduction to Data Mining. Pang-Ning Tan, Michael Steinbach, Vipin Kumar Boston. Pearson Addison Wesley (2006).

Grading scheme:
Assignment 30%
Project 30%
Final Exam 40%

Background needed: COMP271

Available for final year UG students to enroll: Yes but with approval

Minimum CGA required for UG students: None

Course code: COMP5411
Course title: Advanced Computer Graphics
Instructor: Chiew-Lan Tai and Pedro Sander
Room: 3515; 3525
Telephone: 2358-7020; 2358-6983
Email: ;
WWW page: http://course.cse.ust.hk/comp541

Area in which course can be counted: VG

Course description:
Computer Graphics studies the principles of generating and displaying 3D images on the computer display. This course will first cover advanced topics in modeling and processing geometric shapes, and then topics on geometry rendering, lighting, and shading, using latest generation graphics hardware.

Background: COMP271, Linear Algebra, Calculus

Exclusion: CSIT540

Course outline/content (by major topics):
Basics of Computer Graphics
Curves and surfaces (Bezier, b-spline, implicit surfaces)
Mathematical: least squares, PCA, SVD
Discrete differential geometry
Differential methods for shape editing
Space-based deformation
Surface simplification
Surface smoothing
Graphics Processing Unit (GPU)
Programmable Rendering Pipeline (Vertex, Geometry, and Pixel shaders)
Surface lighting and shading
Real-time shadow algorithms
Global illumination
Future trends on GPU computing

Text book:
Dave Shreiner. OpenGL Programming Guide. Seventh Edition. Adisson Wesley. (optional reference book)

Reference books/materials:

Grading scheme:
70% - Four programming projects
30% - Exams

Available for final year UG students to enroll: Yes

Minimum CGA required for UG students: Permission of the instructor

Course code: COMP5621
Course title: Computer Networks
Instructor: Brahim Bensaou
Room: 3537
Telephone: 23587014
Email:
WWW page:

Area in which course can be counted: NT

Course description:
Principles, design and implementation of computer communication networks; network architecture and protocols, OSI reference model and TCP/IP networking architecture; Internet applications and requirements; transport protocols, TCP and UDP; network layer protocols, IP, routing, multicasting and broadcasting; local area networks; data link and physical layer issues; TCP congestion control, quality of service, emerging trends in networking.

Exclusion: COMP362

Course objective:
Upon completion of this course you will have an in depth knowledge about the foundations of current Internet applications, serviced and architecture and will learn about some of the challenges that are defining the future trends in the design of new services and protocols for the Internet.

Textbooks:
   *James Kurose and Keith Ross, Computer Networking: A Top Down Approach, (5th Ed.), Pearson, 2009.

Reference books/materials:
   *A collection of papers and articles provided as a reading list.

Grading scheme:
   *Homework (can be a paper presentation), Mid-term and Final Exam.

Available for final year UG students to enroll: Yes

Minimum CGA required for UG students: Instructor Permission required

Course code: COMP5712
Course title: Introduction to Combinatorial Optimization
Instructor: Sunil Arya
Room: 3509
Telephone: 2358-8769
Email:
WWW page: http://cse.hkust.edu.hk/~arya/

Area in which course can be counted: TH

Course description:
   * An introduction to the basic tools of Combinatorial Optimization.
   * Includes: Linear Programming, Matching, Spanning Trees and Matroids, Dynamic Programming and Basic Graph Algorithms.

Course objective:
   * Upon completion of this course students will have been introduced to many of the most basic tools of combinatorial optimization and will be able to apply them towards designing efficient algorithms in their own research domains.

Textbooks:
   * Combinatorial Optimization: Algorithms and Complexity. Christos H. Papadimitriou and Kenneth Steiglitz, Dover books, 1998.

Grading scheme: Homeworks, midterm and final examination.

Background needed:
   * COMP271 or equivalent + Linear Algebra

Available for final year UG students to enroll: Yes

Minimum CGA required for UG students: Permission of the instructor

Course code: COMP6111A
Course title: Cloud Computing Systems
Instructor: Lin Gu
Room: 3562
Telephone: 2358-6991
Email:
WWW page: http://cse.hkust.edu.hk/~lingu

Area in which course can be counted: ST

Course description:
The commoditization of computers and Internet access capability has given rise to a new computing paradigm called "cloud computing". Integrating gigantic datacenters with geographically distributed user populations, cloud computing can provide high-quality services to individual and corporate users at very low cost. In the meantime, cloud computing also requires that we re-design many components of the computing systems. This course will introduce the design of several Internet-scale systems, and study several technological components of cloud computing systems, including operating systems, storage abstractions, data models, programming frameworks, development utilities, user interfaces, and network design.

Exclusion: COMP660L

Course objective:
This course will study organization of cloud computing systems and survey research problems in this area. This course will re-visit a number of components in computing systems, investigate their new design constraints in the cloud-computing environment, and study the state-of-the-art solutions by surveying both research papers and industrial developments. With a focus on system software design, many topics in cloud computing will be discussed, and a project based on Windows Azure provides opportunities for hands-on experience on solving some real problems in this area.

Course outline/content:
Topics include the system organization in a cloud computing environment, a survey on related works (grid computing, etc.), file systems, parallelization techniques, database semantics, network management, distributed services, and energy efficiency. This course uses Windows Azure as an experimental platform, and will also cover related technologies, such as Amazon EC2 and Google App Engine.

Textbooks: None

Reference books/materials:
Instructor-prepared lecture materials and a collection of research papers.

Grading scheme:
We will meet twice a week for a mixture of lectures and class discussions of assigned readings.
Grades will be based on class participation, reading notes, and a course project. Each student will present one or more assigned papers and lead a class discussion.

Available for final year UG students to enroll: Yes

Minimum CGA required for UG students: Permission of the instructor

Course code: COMP6311A
Course title: Topics in DB: Mobile and Location-Based Search
Instructor: Dik Lun Lee
Room: 3534
Telephone: 2358-7017
Email:
WWW page: http://cse.hkust.edu.hk/~dlee/630/

Area in which course can be counted: DB

Course description:
Information retrieval models, profiling user interests, personalization of search engine results; wireless data dissemination, human trajectory analysis, integration of content, location and social network in search, collaborative search and location recommendation.

Exclusion: COMP630O

Background needed: Background in database.

Course objective:
1. Acquire broad knowledge in data management issues on internet and wireless networks.
2. Develop indepth knowledge in specific topics by carrying out a course project.

Course outline/content (by major topics):
1. Course Overview
2. Information Retrieval and Search Techniques
   - Vector space model, semantic indexing and search, relevance feedback.
   - Web-based search: PageRank, hub and authority.
3. Search Engine Personalization
   - Capturing page preferences: Joachims' methods and Ranking SVM.
   - From page preferences to concept preferences: overcoming click sparsity, capturing user interests with concepts and ontologies.
   - Collaborative filtering and community formation: exploitation of community to enhance search quality, finding the experts and avoiding nuance in seeking recommendations.
   - From queries to tasks: search engine log analysis, understanding what the user is trying to find from sequences of queries.
   - Page summarization: summarizing a page with concept graphs.
4. Location-based Search and Location Prediction
   - What is a location? Capturing location preferences via clickthrough analysis.
   - Trajectory analysis and location recommendation and prediction.
   - Integration of content, trajectory and social network in search and location prediction.
   - Location privacy in mobile and personalized search.
5. Technologies and Performance Concerns in Mobile Data Management
   - Point-to-point versus data broadcast.
   - Wireless data broadcast; indexing and caching issues.
   - Data broadcast for the small.

Textbooks: None

Reference books/materials: Papers from the literature

Grading scheme: Homework assignment, presentation and course project
Class participation: 10%
Homework assignments: 20%
Presentation: 20%
Course project: 50%

Available for final year UG students to enroll: Yes

Minimum CGA required for UG students: Permission of the instructor

Course code: COMP6311B
Course title: Topics in DB: Managing Probabilistic Data in Advanced Database Applications
Instructor: Wilfred Ng
Room: 3505
Telephone: 2358-6979
Email:
WWW page:

* Instructor's approval is needed for registration

Area in which course can be counted: DB

Course description:
As probabilistic data continues to populate in a wide spectrum of new database applications such as RFID and genetic applications, we are facing new challenges that require us to relax many traditional database assumptions and advance existing querying and mining techniques. In addition to relaxing conventional constraints such as functional dependencies and ranking, we study new modeling, querying and mining techniques in many database applications. For example, modeling RFID data and microarraying data are needed to support mining more interesting and useful probabilistic information. Information extracted from independent web sources may also require effective techniques of merging data in a probabilistic manner. In this course, we will discuss in depth the research issues arising from probabilistic data management and its applications.

Exclusion/Background needed: Excellent fundamental database knowledge and basic research techniques.

Course objective:
To gain a better understanding of the current research topics in advanced probabilistic databases and the applications, especially how to store, query, mine probabilistic data in sequence databases, RFID databases, web data, and bioinformatics data. To equip you with the techniques to survey, analyze and criticize research papers, obtain hands-on experience on database projects and participate research with other students.

Course outline/content (by major topics):
1. Basics probabilistic data models.
2. Ranking in probabilistic databases.
3. Data Dependencies in probabilistic relational databases.
4. Mining probabilistic sequence databases.
5. Probabilistic micro arraying data mining.
6. Probabilistic convex hull query.
7. Probabilistic RFID Data modelling.
8. Probabilistic Data and Web searching and mining.

Textbooks: No need.

Reference books/materials:
   * Database journals such as IEEE TKDE, ACM TODS and Information Systems.
   * Conference Proceedings such as SIGMOD, VLDB, ICDE, PODS, EDBT, KDD, ICDM.
   * Electronic Resources:
     1. ACM Digital Library: http://portal.acm.org/dl.cfm
     2. CiteSeer Computer Science: http://citeseer.ist.psu.edu/
     3. ACM Special Interest Group on Management of Data (SIGMOD): http://www.acm.org/sigmod
     4. ACM Special Interest Group on Information Retrieval (SIGIR): http://www.acm.org/sigir
     5. Very Large Data Bases (VLDB): http://www.vldb.org/
     6. DBLP: http://www.informatik.uni-trier.de/~ley/db/index.html
     7. IEEE Computer Society: http://www.computer.org/
     8. IEEE Explore: http://ieeexplore.ieee.org/

Grading scheme:
No Final examination. Paper Presentation (30%), Survey Assignment (30%), Individual Research Report (40%).

Available for final year UG students to enroll: Subject to instructor's approval.

Minimum CGA required for UG students: A or above.

Please visit https://www.ab.ust.hk/wcr/cr_class_staf_main.htm for the timetable and quota.

Archive of past courses

Last modified by Yongxin Tong on 2011/08/24.