Machine Learning: An Undergraduate Engineering Course

Abstract In today's quickly changing world, staying up-to-date is a recipe for success. This is particularly true in the Electrical and Computer Engineering (ECE) field. While the main concepts in circuits and programming remain unchanged, the tools and the applications are changing at a very fast pace. As a result, curriculum committees within ECE programs are continuously striving to update their curricula to prepare their graduates for success in both the industrial and the research worlds. Machine learning (ML) is an ECE area that continues to gain an increasing popularity in the ECE industries. Consequently, it is becoming a staple on the list of "highly-demanded skills" for many of the ECE employers. In order to fulfill this workforce need and help increase the employability of our graduates, an undergraduate machine learning course has been developed and will be taught in Spring 2022. The course will be offered as a 3 credit upper level technical elective for Electrical Engineering, Computer Engineering, Mechanical Engineering, and Computer Science students. Several challenges were encountered when developing this course. First of all, considering the myriad of topics that can fit under the ML umbrella, it was necessary to find an adequate breadth vs depth balance to design a course that fits within the time constraints of an academic calendar. Also, considering the undergraduate nature of the course, some of the high-complexity mathematical concepts had to be simplified without sacrificing the rigor of the course. Another challenge was finding the right balance between covering the conceptual background behind the various ML methods as opposed to exclusively teaching the practical tools and techniques that allows the students to find solutions without looking "under the hood". Finally, as numerous computing platforms can be used to implement ML systems, a thorough investigation was executed to identify the most adequate platform to use in the course. This paper will share the ins and outs of the newly developed course. It will discuss the appropriate topics to be included and the adequate amount of depth for an undergraduate audience. It will also discuss the selected computing platform with the rationale for all the decisions.