6.036: Introduction to Machine Learning
If you are a current student, please Log In for full access to the web site.
Note that this link will take you to an external site (https://cas.mitx.mit.edu) to authenticate, and then you will be redirected back to this page.
Instructions for taking the Final Exam (Monday 03/24)
If you are taking the exam at the regular time (1:30PM - 5:30PM) and don’t have an accommodation.
At 1:30 PM:
- Login to our course site on canvas (your account should already be set up: please use your MIT kerberos account).
- Click on Gradescope (among the options on the left).
- Click on the relevant assignment "Final Exam PDF upload" or "Final Exam IMAGE upload" depending on whether you're uploading a PDF or images.
- An upload dialog box should open.
- Click on the link to download the PDF for the exam. Yes, the link to the exam PDF is included in the upload dialog box.
If possible, either print the pdf and write on it, or write on it using a tablet.
If not, then write your answers, clearly separated and marked, on blank paper. Please start a new page for each question.
By 5:30 PM
- In the Gradescope upload dialog, follow the instructions for your type of upload.
- If you are uploading images, you will be asked to indicate which image has the answer to each question. You need to Submit after you have done that.
- Go back to Gradescope (options on the left) and verify that your assignment Status says Submitted.
- It is your responsibility to ensure that all of your pages have been uploaded correctly.
- If you have a logistical problem, mail your solution as one file (PDF or ZIP) to email@example.com and make a private post on Piazza.
If you are taking the exam at a different time or have an accommodation, you will not be able to get your exam or upload your answer via Gradescope. Follow the instructions that you received by email from firstname.lastname@example.org.
- This is an open-book exam. You may use any materials you want (electronic or otherwise, including notes, calculators, the 6.036 online material, Python, and Wikipedia) during the exam, but you are not allowed to converse with other humans (including through text message, email, etc.) from the time you start the exam until 24 hours afterward.
- It was designed to be a 3-hour exam, but you have an extra hour to help account for upload/download time.
- The problems are not necessarily in any order of difficulty.
- Record all your answers in the places provided. If you run out of room for an answer, indicate that you are continuing your answer, write on a blank page and mark clearly what question is being continued, and append a scan of the page to your submitted file.
- If a question seems vague or under-specified to you, make an assumption, write it down, and solve the problem given your assumption.
- We will not be answering questions about the exam via email or Piazza. An exception is if you are very sure there is a significant error, in which case, please make a private Piazza post. Any information we provide in response will be emailed to the entire class.
Lucky Week 13 Announcements
It's week 13!
Deadline for submitting all course assignments is 05/20 at 11.59pm ET. Automatic extensions do not apply beyond this date.
Lecture 13: Tuesday 5/11 at 9:30am EDT on Unsupervised learning: Clustering.
Exercises 13: due on 5/11 at 9am EDT before the lecture.
Lab 13: Thursday, 5/13 & Friday, 5/14; Checkoffs are due lab this week--not on the following Monday.
No more new homework!
Homework 11: due on Monday, 5/10 at 11PM EDT (extended due to MIT student holiday).
Homework 12: due on Friday, 5/14 at 11PM EDT.
If you cannot take the final at the scheduled time (1:30PM - 4:30PM) due to time-zone or internship or other substantial reason and you don't already have a conflict-exam message from the registrar, please make a private Piazza post explaining your situation and need for the 9PM alternate start time.
The fourth Nanoquiz (also called Week 11 Nanoquiz) will be available at this url at 12.01am ET on Wednesday (04/28/2021). You can also get to the Nanoquiz from the course website (introml.odl.mit.edu) and then from under Week 11.
You will have 48 hours to start the Nanoquiz and 60 minutes to complete it. Please do not communicate with anyone during this time and do not post on Piazza. Once you visit the Nanoquiz page (link above), you will have to manually activate it (the instructions will be on the page itself) and once activated, your timer will begin. You will not be able to make any changes once the timer runs out.
Remember to actually hit 'SUBMIT' for each of the questions! If you do not, you won't get your points! Once you hit Submit, you should see a 'Grade not available' text appear. Unless you see it, your submission will not be recorded and cannot be graded.
Week 9 announcements (04/12 - 04/16)
Lecture 09: Tuesday 4/13 at 9:30am EDT on learning RNNs.
Exercises 09: due on 4/13 at 9am EDT before the lecture.
Lab 09: Thursday, 4/15 & Friday, 4/16; due Wednesday 4/21 at 9pm EDT.
Homework 08: due on Friday 4/16 at 11pm EDT.
Fourth Nanoquiz (nq09): opens at 12.01am EDT on Wednesday 4/14 and stays open until 11.59pm EDT on Thursday 4/15. You will have 60 mins to complete it. Nanoquiz prep materials have been updated with more material on CNNs and MDPs.
Please submit an NQ3 rebuttal, if you plan to do so, before Saturday April 9.
Note also that in Week 10, there is no lecture, no reading, no exercises, no new HW. Just a cool lab!
Week 5 announcements (03/15 - 03/19)
Coming up next week:
Monday: Get your checkoff on Lab 04 if you didn’t get it in lab.
Tuesday: Lecture 05 on Neural Networks.
Tuesday: Exercise2 05 due by 9AM EDT!.
Wednesday: Homework 04 due by 11PM EDT.
Wednesday-Thursday: Second Nanoquiz (see details below).
VideoQA for week 3 including NanoQuiz 1 will be available.
All 6.862 project proposal meetings are next week Tuesday 03/16 - Friday 03/19.
Second Nanoquiz (Wednesday-Thursday)
It will be available for 48 hours from 00:00 EDT on Wed 3/17 (Midnight Tue night in Boston) through 23:59 EDT on Thur 3/18 (Midnight Thu night in Boston). It will be open for you to work on for 1 hour within that window. It will cover material from weeks 1 through 4 at the level of depth of old midterm questions.
Remember to hit 'SUBMIT' for each of the Nanoquiz questions! If you do not, you won't get your points! Once you hit Submit, you should see a ‘Grade not available' text appear. Unless you see it, your submission will not be recorded.
You can find relevant practice problems from old midterms on the Nanoquiz Information and Materials page.
The first Nanoquiz (also called Week 3 Nanoquiz) will be available at this url at 12.01am on Wednesday (03/03/2021). You can also get to the Nanoquiz from the course website (introml.odl.mit.edu) and then from under Week 3.
You will have 24 hours to start the Nanoquiz and 60 minutes to complete it. Please do not communicate with anyone during this time and do not post on Piazza. Once you visit the Nanoquiz page (link above), you will have to manually activate it (the instructions will be on the page itself) and once activated, your timer will begin. You will not be able to make any changes once the timer runs out.
Nanoquiz grades will be visible to you early next week.
Announcements: Week 3 (Mar 2-5)
Here are some announcements for the week (Week # 3).
Lecture 03 on Linear and Logistic Regression will be given on Tuesday, 3.2 at 9:30am.
Exercise 03 is due before lecture at 9am.
Lab 03 will be as usual on Thursday, 3.4 or Friday, 3.5, and is due the following Monday 3.8 11pm.
Homework 03 will be out Thursday 3.4, and due Wednesday 3.10 at 11pm.
VideoQA's are recorded video solutions of all exercise and homework questions. The first VideoQA's, for week 1, will be available Monday 3.1.
The first nanoquiz is next week and will be open for any 1-hour period on Wednesday 3.3. See the Nanoquiz Information and Materials page for more information.
Be sure to watch the new VideoQA's, and good luck on your nanoquiz!
Announcements: Week 2 (Feb 23-26)
We hope you’ve had a good first week of 6.036! Here are some announcements for the week (Week # 2).
New students: if you have registered for 6.036 after the first day of classes and are catching up on any of the first week assignments, please read week 1 announcements on the course website (introml.odl.mit.eduLinks to an external site.).
Missed lab: If you missed Lab 1 during Week 1 do not worry, there are required submission components and you should work those to complete them, and get a checkoff for required discussion components in an online office hour. Labs are normally completed during your live lab/MLyPod section time, but if you do not finish them the checkoff can be completed without late penalties usually by the following Monday at 9pm.
Grade of previous activities: A page is available showing your scores on previous assignments, including any lateness that applies. You can find this under Progress at the top of the 6.036 home page.
Lab score: Please use the progress tab to make sure you entered answers to all lab questions to get full 1.0 credit on the lab.
Attendance: After week 1, attendance at your lab section is mandatory and constitutes 5% of the grade. You may change your lab section (during the next two weeks) by moving into any pod with space by going to introml.odl.mit.edu > Lab/Pod Self Schedule.
Exercises: out now, due Tuesday 2.23 at 9am EST (in general, out on Thursday and due on Tuesdays with a lecture).
Homework: out now, due Wednesday 2.24 at 11pm EST (in general, out on Thursday and due on the next Wednesday evening).
Office hours: Just join the help queue (available from the Information tab) and you’ll get assigned to a staff member and then receive information on what Zoom room to join. The schedule is:
For more information, see Office Hours details.
Announcements: Week 1 (Feb 16-19)
Exercises: out Friday 2.19, due Tuesday 2.23 at 9am EST
Lab: Thursday, 2.18 & Friday, 2.19, due Monday 2.22 at 11pm EST
Homework: out Friday 2.19, due Wednesday 2.24 at 11pm EST
Lecture: A live lecture via zoom is available at the normal lecture time, Tuesdays 9:30am to 11:00am EST, with an opportunity for online Q&A these will be recorded and available asynchronously afterward. In addition, prior semester video recordings will be available asynchronously.
Exercises: Online exercises will be due each Tuesday by 9:00am EST, based on weekly online readings.
Labs: Each student will match (be assigned) to a weekly 1.5 hour lab session; students will attend the same session/time each week by zoom. Multiple session times will be available across Thursday and Friday, including some to accommodate for students in different time zones. The lab session will be synchronous and required (i.e., live participation will be part of the grade). Instructions about signing up for Lab sections are provided in the announcement immediately below this one.
Homeworks: Online weekly homeworks are due each Wednesday at 11:00pm EST
Exams: There will be a final and 5 nano-quizzes.
Office hours: Staffed online office hours (with a help queue and videochat connection) will be available during standing scheduled times each week, for individual help and for completing lab checkoffs. In addition to promote student collaboration and community there will be a virtual student room open 24/7 and sampled by LA’s during OH’s.
Lab Section/Pod Scheduling (Feb 15, 2021)
As promised, here is the information about choosing your Lab sections. Please read carefully and pay special attention to the deadline noted below:
Please navigate to the schedule survey provide your scheduling preferences. This schedule survey will remain open until midnight (Eastern) on Tuesday (02/16/2021). We will then assign everyone to Lab sections/pods based on the survey. Your section assignments should be live by Noon (Eastern) on Wednesday (02/17/2021). Your first Lab will be on Thursday (02/18/2021) or Friday (02/19/2021). You should also read through the "Lab Sections" section on the Basic Information page describing how this remote version of the subject will operate this Spring.
After Noon (Eastern) on Wednesday (02/17/2021), you should be able to view your assigned Lab Section/Pod. We will send an update about how you can view this information on Wednesday. For now, please fill out the schedule survey described above (1).
After Noon (Eastern) on Wednesday, if you would like to switch to a different section, you will have the opportunity to self-switch to any other open Sections/Pods with capacity. The self-switching option will remain open during the first two weeks of the semester. Instructions about how to do so will be shared on Wednesday. For now, please fill out the schedule survey described above (1).
We encourage you to avoid switching sections frequently. The labs are designed to enable dedicated staff members (instructor + TA + LA) to get to know you during the semester. If you switch too frequently, you may deny yourself (and the staff) the opportunity to get to know each other better.
After the first two weeks, if you would like a permanent switch to a different lab section, please email email@example.com. We will support temporary one-off switches on a case-by-case basis but you will need approval before a switch is allowed. You cannot attend a different Lab Section/Pod without permission. There are no exceptions.
Please note that Listeners will not be assigned to any Lab Sections/Pods.
Important Information (Feb 11, 2021)
The course (6.036/6.862) officially starts with the first lecture on Tuesday February 16, 2021 (9.30-11.00am). We expect to make the course assignments available to students on a weekly-basis. At the moment, course assignments are not accessible to the students. Please refer to the updated course calendar for more information.
There are no exercises or homework due in the first course-week (the first exercises are due on Tuesday February 23, 2021 and the first homework is due on Wednesday February 24, 2021). They will become accessible to students after the first Lab.
There will be a Lab on Thursday/Friday of the first course-week.
There is no Nanoquiz scheduled during the first two weeks of the course.
Preparation and Self Assessment
You cannot understand machine learning without understanding vectors, dot products, matrices and partial derivatives well. We also expect you to feel very comfortable with writing and, importantly, debugging Python programs. To help you judge your preparation, 6.036 students are encouraged to work through the (ungraded) Readiness Self-Assessment. If you find yourself struggling with that material, you might consider taking additional linear algebra and python programming subjects before 6.036, and/or review corresponding material on your own.
For example, you should feel comfortable with roughly the first 6 weeks of this 18.02 material. You should be able to do 18.02 practice exams 1 and 2 (it’s fine if you need to use some reference material or look things up) from 18.02. Some additional exposure to linear algebra (linear independence) and probability (conditional probability, expectations) is helpful but can be picked up along the way.
Cross Registered Students
If you are cross-registering, please do so as early as possible. In order to access the 6.036 site, you will need to have an MIT @mit.edu email address and identity, so it is important to register as early as possible. You can contact us at firstname.lastname@example.org if your cross-registration is delayed.
Online structure for Spring 2021
We will have a mixture of asynchronous and required synchronous components in the Spring. The following are our current plans (these may evolve as we get closer to the start of the semester).
A live lecture via zoom will be available at the normal lecture time, Tuesdays 9:30am to 11:00am Eastern, with an opportunity for online Q&A these will be recorded and available asynchronously afterward. In addition, prior semester video recordings will be available asynchronously. Online exercises will be due each Tuesday by 9:00am Eastern, based on weekly online readings.
Each student will match (be assigned) to a weekly 1.5 hour lab session; students will attend the same session/time each week by zoom. Multiple session times will be available across Thursday and Friday, including some to accommodate for students in different time zones. The lab session will be synchronous and required (i.e., live participation will be part of the grade). We are especially excited about using the lab to engage students with each other in teams and with staff, to explore fundamental concepts in advance of individual work in the homeworks.
In order to match you to your lab session, we will ask you to fill a form which will be made available on this website. Stay tuned for more information on Lab section sign-ups. Once you are assigned to a Lab section you will have the opportunity after that to switch into a different lab section time, if there are remaining open slots in the other lab section time. Note that if you are no longer planning to register for 6.036 or 6.862, it would still be helpful for you to go to the survey and note this as the response for the first question, to avoid us assigning you to a section.
Listeners: students registering as Listener will not be able to participate in the live lab/MLyPod sessions, due to capacity limitations. Listeners have access to the lecture, and all of the other online content.
home 5. There will be a Final Exam but no Midterm. We will also have bi-weekly nanoquizes which will include all course-material covered up to the prior week. We expect this to be open-book but with no collaboration allowed. Expanded time for the nanoquiz is planned (60 minutes), with flexible time window for taking it (within 24 hour window before the weekly lab session).
Online weekly homeworks will be due each Wednesday at 10:00pm Eastern.
Staffed online office hours (with a help queue and videochat connection) will be available during standing scheduled times each week, for individual help and for completing lab checkoffs.
6.862: The graduate version coincides with 6.036 in lectures, problem sets, and exams. In addition, 6.862 includes a semester-long class project and online meetings for the project kickoff, proposal, milestone, report, and presentation. Canvas handles the project part. 6.862 is for non-EECS students. There will be no cap on enrollment in the Spring. Grading for 6.862 is split evenly (50/50) between the 6.862 project and 6.036.