A pdf version of the syllabus is here

The Spring 2025 edition welcomes students from multiple departments: the traditional STEM departments of ECE, other engineering, Computer Science, Astronomy, Math, SOEST and others, but also from other departments such as (but not only limited to) economics, business and urban planning. The course centers around a few topics that are carefully selected, but which give students an opportunity to spin off and pursue an angle that is best matched with their background and research agendas.

Basic familiarity with statistics, some probability and linear algebra, and programming is required, supplemented with a willingness to learn fundamentals. Ideally, students should have had some exposure to using machine learning/AI (even if it is only YouTube videos or playing with them as black boxes). The goal is for all students to pick up the central tenets of this field in a way that can guide them to use AI/machine learning in a sophisticated and nuanced fashion. You can take this course even if you have taken other ICS courses on the topic, or if you have taken the large language models course in the ECE department.

This course will also have mentors. Mentors are chosen from students who have taken prior editions of the course. I hope you will agree to mentor students down the line.

The course will be divided into several modules. There is no single text for the course, but there are several excellent resources that may serve you well beyond the course. Here is an (incomplete) list of some of the sources we will use, almost all have legal online versions. That said, some of material in the class is quite new and ongoing research, so I am not aware of textbooks, only technical papers.

1. Dive into Deep Learning’’: online book by Zhang et. al. available here
2. Understanding machine learning: From Theory to Algorithms by Shai Shalev-Shwartz and Shai Ben-David, available here
3. Patterns, Predictions and Actions’’ by Moritz Hardt and Benjamin Recht, available here
4. Learning theory from first principles by Francis Bach, available here
5. Neural Network Learning: Theoretical Foundation by Martin Anthony and Peter Bartlett (an old but really good book). I don’t know if there is an online version.

Modules will have assignments. Genera