COMP 411: Tutorial on Machine Learning
Fall 2021
         Home Schedule Policies Resources         

Tutorial info

  • Professor: Victoria Manfredi, vumanfredi [at], Exley 627, 860-685-2194

  • Meetings: Th/Fr afternoon

  • In-person office hours (in my office): Mon 3:30-5p, Tu 4:30-6p, Wed 4:30-6p, and by appointment.

  • Virtual office hours (on Zoom): Email me to set-up a time to meet, sometime within my in-person office hours if possible.

  • Announcements and online discussion: We'll use Google Classroom. I will add you, but you should double-check that you've been added and receive announcements, etc.


This tutorial provides an introduction to the field of machine learning: the first half of the tutorial will focus on specific regression and classification techniques while the second half of the tutorial will focus on deep neural networks and reinforcement learning. The tutorial readings will be taken from several machine learning textbooks, and will be complemented by a set of slides. There will be weekly meetings to discuss the material and work on programming projects.


This class has COMP 211, COMP 212, and MATH 228 as pre-requisites. We will cover concepts from probability and linear algebra as needed to ensure the course is self-contained. Programming will be done in python.


Your grade will be based on approximately 10 homework assignments (80%), your comments on my slides and homework assignments (10%), and writing up questions and answers for one homework in latex (10%). Assignments will be posted on the schedule. All work must be submitted electronically. Grades and feedback on code submissions will be returned as paper, with written work and code printed out with comments written on it directly. I will email you your grades periodically to double-check that the information that I have is consistent with what you have. It is your responsibility to check your grades and feedback and report any issues promptly. It is always possible for mistakes to happen in recording scores, especially in larger classes. If I do not hear about a problem, it will not be fixed.