2024 Faculty Courses School of Computing Department of Computer Science Graduate major in Artificial Intelligence
Sparse Signal Processing and Optimization
- Academic unit or major
- Graduate major in Artificial Intelligence
- Instructor(s)
- Shunsuke Ono
- Class Format
- Lecture (HyFlex)
- Media-enhanced courses
- -
- Day of week/Period
(Classrooms) - 3-4 Mon / 3-4 Thu
- Class
- -
- Course Code
- ART.T465
- Number of credits
- 200
- Course offered
- 2024
- Offered quarter
- 3Q
- Syllabus updated
- Mar 14, 2025
- Language
- English
Syllabus
Course overview and goals
Summary: This lecture is on signal and information processing techniques based on sparsity. Related optimization techniques are also introduced.
Aim: The aim of this lecture is to understand why sparsity is important, how to model sparsity, why optimization techniques (especially nonsmooth optimization) are required, and what applications exist. Also, students will implement some sparse signal processing techniques by MATLAB/Python. In addition, students will survey papers related to sparse signal processing and optimization to see cutting-edge research.
Course description and aims
Students can explain and implement sparse signal processing and related optimization techniques.
Keywords
Sparse and low-rank signal processing, nonsmooth optimization, compressed sensing, signal recovery
Competencies
- Specialist skills
- Intercultural skills
- Communication skills
- Critical thinking skills
- Practical and/or problem-solving skills
Class flow
We alternate lectures and exercises (MATLAB/Python). Every student will survey related papers and give a paper presentation at the last two lectures.
Course schedule/Objectives
| Course schedule | Objectives | |
|---|---|---|
| Class 1 | Introduction | Understand the outline and aim of the lecture. |
| Class 2 | [Lecture] Mathematical ingredients | Understand basic mathematical tools for sparse signal processing. |
| Class 3 | [Exercise] Mathematical ingredients | Implement basic mathematical tools for sparse signal processing. |
| Class 4 | [Lecture] Proximal gradient method and sparse signal estimation | Understand the proximal gradient method and its application to sparse signal estimation. |
| Class 5 | [Exercise] Proximal gradient method and sparse signal estimation | Implement sparse signal estimation by the proximal gradient method. |
| Class 6 | [Lecture] Alternating direction method of multipliers and robust principal component analysis | Understand the alternating direction method of multipliers and its application to robust principal component analysis. |
| Class 7 | [Exercise] Alternating direction method of multipliers and robust principal component analysis | Implement robust principal component analysis by the alternating direction method of multipliers. |
| Class 8 | [Lecture] Primal-dual proximal splitting method and image restoration | Understand the primal-dual proximal splitting method and its application to image restoration. |
| Class 9 | [Exercise] Primal-dual proximal splitting method and image restoration | Implement image restoration by the primal-dual proximal splitting method. |
| Class 10 | [Lecture] Advanced applications | Understand advanced applications. |
| Class 11 | [Exercise] Advanced applications | Implement advanced applications. |
| Class 12 | Survey and reading | How to read and introduce papers. |
| Class 13 | Paper presentation I | Introduce and discuss papers on sparse signal processing and optimization. |
| Class 14 | Paper presentation II | Introduce and discuss papers on sparse signal processing and optimization. |
Study advice (preparation and review)
To enhance effective learning, students are encouraged to spend approximately 100 minutes preparing for class and another 100 minutes reviewing class content afterwards (including assignments) for each class.
They should do so by referring to textbooks and other course material.
Textbook(s)
No need to buy textbooks
Reference books, course materials, etc.
Materials for the course will be provided online.
Evaluation methods and criteria
Course marks are based on exercises (source code, 50%) and paper presentation (slides and Q&A, 50%).
Related courses
- MCS.T402 ： Mathematical Optimization: Theory and Algorithms
- ART.T458 ： Advanced Machine Learning
Prerequisites
Required: Linear algebra, differential and integral analysis, probability theory, statistics, and programming experience on MATLAB/Python
Recommended: Functional analysis, numerical calculation
Other
Every student must bring a laptop computer with MATLAB/Python for exercises. You can choose MATLAB or Python but sample code in slides will be written by MATLAB.