From Advanced Digital Signal Processing to Machine Learning

Contribution: A modular three-credit senior undergraduate or graduate course that includes both advanced digital signal processing (DSP) and machine learning (ML) is introduced. This project-based course is featured at exploiting the connections between the two popular areas. It presents the advantages and limitations of each area to the students with the help of various specially designed course projects. Background: At most schools, DSP and ML are taught in two separate courses although the two have many similarities in both theorems and applications. It's partially due to the fact that the two areas are largely supported by two research communities in the literature. Intended Outcomes: Students understand the fundamentals of the DSP and the ML from both theorems and applications perspectives with a clear picture of the advantages and limitations in the state-of-the-art. Set up initial motivations and inspire the students to the future research activities. Application Design: An approach that concentrated content and project work on the modular domain knowledge was taken. The authors started with the probability and statistics, followed by a series of carefully selected topics from each area that can be combined in a systematic way, from theorems to applications, from algorithms to implementations. An oral presentation session was included to emphasize independent analysis and professional skills development in the student selected project. Findings: Based on the course survey, the students reported positive feedbacks on the newly structured course. It showed that this course can benefit from the strategic introduction of desirable technology in the curriculum efficiently. From the theorems perspective, students reported the domain knowledge of the two areas. From the applications perspective, students were exposed to the state-of-the-art tools in two disciplines, career preparation, and in-class engagement. A direct assessment showed a promising result.