SDSC6011 Course Information

#sdsc6011 #course information

English / 中文

Course Overview

Course Code: SDSC6011

Course Name: Optimization for Data Science

Semester: Fall, 2025

Instructor: Professor Jun Wang

Email: jwang.cs@cityu.edu.hk

Office: Room G7361, Yeung Kin Man Academic Building

Tel: 34429701

  • Lecture Time: Tuesday 19:00-20:50

  • Tutorial Time: Tuesday 20:00-21:50

  • Teaching Mode: Face-to-face

  • Location: Lecture Theatre 5 (LT5), Yeung Kin Man Academic Building

  • Teaching Assistants:

    • Lin Li (linli82-c@my.cityu.edu.hk), LAU-16-290 Seat 16
    • Chenghao Liu (chenghliu9-c@my.cityu.edu.hk), LAU-16-290 Seat 35
    • Tianze Zhang (tzhang585-c@my.cityu.edu.hk), No office available

Textbook:

Convex Optimization (3rd edition), Stephen Boyd and Lieven Vandenberghe, Cambridge University Press.

Reference Books:

  • Lectures on Modern Convex Optimization: Analysis, Algorithms, and Engineering Applications, Aharon Ben-Tal, Arkadi Nemirovski

  • Linear and Nonlinear Programming (3rd edition), David G. Luenberger and Yinyu Ye

Course Outline:

This course offers an introduction to optimization methods with applications in data science. We will introduce the theoretical foundation and fundamental algorithms for optimization, covering advanced methods for large-scale problems in data science and machine learning. Content includes linear and nonlinear programming, conic programming, convex analysis, Lagrangian duality theory, augmented Lagrangian methods, and stochastic gradient descent. Students will implement algorithms in a programming language and test them on realistic datasets.

Grading Policy

Component Weight Details
Homework 20% Individual assignments. Late submissions may be penalized.
Midterm Exam 40% Closed-book; one A4 cheat sheet allowed.
Final Exam 40% Comprehensive coverage; closed-book; one A4 cheat sheet allowed.

Course Objectives

This course provides students with:

  1. Ability to understand basic concepts of practical optimization problems.

  2. Ability to formulate scientific and engineering problems as optimization tasks.

  3. Ability to solve linear/nonlinear programming problems using analytic and numeric methods.

  4. Ability to apply optimization approaches to science, engineering, and management domains.

Teaching/Learning Expectations

Students must attend classes regularly and maintain quiet during lectures. Instructors will not cancel sessions without official approval. Collaborative discussion is encouraged for idea generation, but final work must reflect individual originality.

Academic Honesty

Plagiarism is strictly prohibited and may result in grade penalties or suspension. Students may discuss homework preliminarily, but submissions must demonstrate personal effort and originality.