ELEC97088 (EE4-66) Topics in Large Dimensional Signal Processing
Lecturer(s): Dr Wei Dai
This module is designed for students to gain skills on modeling and algorithm-design in processing and analyzing large dimensional data, and to expose them to various application domains including image/video processing, machine learning, collaborative filtering, social network, financial data analysis, etc.
Upon successful completion of this module, students will be able to:
1. Present ideas, modelling, theory, and algorithm designs for large dimensinoal data processing, with the focus on finding sparse structures hidden in the data for dimension reduction;
2. Demonstrate that basic tools in linear algebra, optimization and statistics can be developed to solve real-world problems involving large dimensional data.
3. Gain skills to work with real-world applications including computational imaging, online recommendation, machine learning, etc.
1. Introduction to large dimensional data processing: the key challenge.
2. Mathematical tools for large dimensional data processing
** Matrix analysis
** Optimization: Convexity, duality, and algorithms
3. Formulations to explore sparsity in the data
** Linear inverse problems
** Sparse linear inverse problems
** Sparse bilinear inverse problems
4. Selected real-world applications
** Google page rank, compressed sensing, denoising, robust classification, Netflix problem, blind deconvolution, super-resolution, etc.
Exam Duration: N/A
Coursework contribution: 100%
Closed or Open Book (end of year exam): N/A
Coursework involves paper study and 6 minute presentation. The detailed format and procedure will be announced in lectures.
The paper study is designed to encourage students to go beyond the taught materials and cultivate a good taste about important techniques
and applications. The final presentations will help students largely broaden their views of the topic, witness how their peers use their
judgement to choose a sub-topic to study, and get exposed to critical thinking of others.
Oral Exam Required (as final assessment): N/A
Prerequisite module(s): None required
Course Homepage: unavailable
Please see Module Reading list