Basic Information About 6.390

You are not logged in.

Please Log In for full access to the web site.
Note that this link will take you to an external site (https://shimmer.mit.edu) to authenticate, and then you will be redirected back to this page.

1) Course Overview §

6.390 introduces the principles and algorithms of machine learning from an optimization perspective. Topics include linear and non-linear models for supervised, unsupervised, and reinforcement learning, with a focus on gradient-based methods and neural-network architectures. Enrollment may be limited.

2) Prerequisites §

Concretely, things we expect you to know (we use these constantly, but don’t teach them explicitly):

2.1) Programming §

  • Intermediate Python, including the notion of classes.
  • Exposure to algorithms – ability to understand & discuss pseudo-code, and implement in Python.

2.2) Linear Algebra §

  • Fundamental matrix concepts and manipulations, e.g., rank, multiplication, and inverse.
  • Points and planes in high-dimensional space.
  • Basic matrix calculus, e.g., gradients.

6.1010 or 6.1210 can serve as the programming prerequisite. 18.06, 18.C06, 18.03, or 18.700 can serve as the linear algebra prerequisite.

(For each of these courses above, a link points to a representative syllabus from some past semesters, for reference.)

3) Course Components§

3.1) Lectures§

Lectures focus to anchor the upcoming week's discussion, overview the technical contents, and tie together the high-level motivations, concepts, and stories. Along with notes and the exercises, lectures prepare students for the upcoming Friday recitations and Tuesday labs.

Lectures will be held class-wide, in Room 10-250, Thursdays 11am-12:30pm. Recordings will be made available shortly after live sessions.

3.2) Exercises§

Online exercises are typically released on Thursday (available and completed through the course website) 1pm, and will be due the following Tuesday by 9am.

The intention is for you to read the notes and/or viewed the Thursday lectures, and do these exercises, so as to maximize the value of your participation in the upcoming recitation and lab, and to begin learning the material in advance of the next lab and homework.

3.3) Recitations§

Recitations reinforce problem-solving skills through examples and pen-and-paper problems. Recitations are held in four sections, all in 32-044: Fridays 10-11am, 11am-12pm, 1-2pm, and 2-3pm.

You're welcome to attend any recitation section that fits your schedule, subject to room capacity.

3.4) Labs§

The Tuesday section meeting will be a Lab assignment that you work through with a student partner and get in-lab checkoffs on.

Each student must attend a weekly 2 hour lab session on Tuesdays. The lab session will be synchronous. We will be using the lab to engage students with each other in small teams (typically two to three students per team) and with staff, to explore fundamental concepts in advance of individual work in the homeworks.

Typically, each lab will require a few "checkoff"s --a brief discussion with a staff member on the topic of the assigned problems. Labs are due in session. However, if the first checkoff is completed by the end of the lab section meeting, then the rest of the checkoffs can be completed, without penalty, by 9pm the Monday after the lab section in office hours.

If you are sick, please do not attend lab; For illness or personal situations, see the guidelines below.

Section Time Room
1 9am-11am 34-501
2 11am-1pm 34-501
3 11am-1pm 32-044
4 1pm-3pm 34-501
5 1pm-3pm 32-044
6 3pm-5pm 34-501

3.5) Homeworks§

Homework is generally released each Friday at 9am Eastern, and is due online (through the course website) the following Wednesday (12 days later) at 11:59pm Eastern.

3.6) Midterm and Final Exam§

There are two midterm exams and one final exam.

The first Midterm Exam is scheduled on Wednesday, October 8, 7:30pm to 9pm.

The second Midterm Exam is scheduled on Wednesday, November 12, 7:30pm to 9pm.

We will also have a Final Exam. The exact time is to be scheduled by the Registrar and details will be annouced around the third week of the term. All three exams are in-person and written exams.

4) Getting help on 6.390§

Please follow this guideline when asking for help (and note that the best way to get help depends on the kind of question you have).

5) Illness and personal issues§

Please refer to the Grading page for more info.

6) Listeners§

Due to capacity and other constraints, we will not accept Listener registrants in 6.390 this semester. We are sorry about that, and can offer, as an alternative, the complete course material from three years ago, including lectures, readings, and online homework.