Lectures: Tuesday and Thursday, 1.25-2.40pm, Baker Lab 200
Course overview:
The course provides an introduction to machine learning, focusing on supervised learning
and its theoretical foundations. Topics include regularized linear models, boosting, kernels, deep networks,
generative models, online learning, and ethical questions arising in ML applications.
Prerequisites (not corequisites):
Probability theory (e.g., BTRY 3080, CS 2800, ECON 3130, ENGRD 2700, MATH 4710) and linear algebra
(e.g., MATH 2210, MATH 2310, MATH 2940), single-variable calculus (e.g., MATH 1110, MATH 1920) and programming
proficiency (e.g., CS 2110).
(Please see the FAQ on corequisites for more information.)
Course logistics:
For enrolled students the companion Canvas page serves as a
hub for access to Ed (the course forum), Vocareum
(for course projects), Gradescope (for HWs), and paper
comprehension quizzes. If you are enrolled in the course you should automatically have access to the site. Please
let
us know if you are unable to access it.
Homework, projects, and exams
Your grade in this course is comprised of three components: homework, exams, and projects. Please also read
through the given references in concert with the lectures.
Homework
There will be a number of homework assignments throughout the course, typically made available roughly one to
two weeks before the due date. The homework primarily focuses on theoretical aspects of the material and is
intended to provide preparation for the exams. Homework may be completed in groups of up to four. The
assignments themselves will be made available via Gradescope (through Canvas).
You are allowed two slip days per homework assignment.
Projects
To provide hands-on learning with the methods discussed in class there are a number of programming projects
throughout the course. The projects may be completed individually or in a group of two. They are accessed,
submitted, and graded using Vocareum.
You are allowed two slip days per project (same as for homework).
Paper comprehension (CS 5780)
Students enrolled in the graduate version of the course (i.e., in CS 5780) are required to read the assigned
research paper(s) and complete the associated online quiz. Paper(s) will be assigned roughly once every two to
three weeks. This is an individual component (not to be done in groups). The quizzes will be
made available on Canvas.
You are allowed two slip days per project (same as for homework).
Exams
There will be two exams for this class (to be completed individually, not in a group), an evening
prelim and a final exam. The location and time for the final is TBD; we will update the date
below once the registrar finalizes it.
Prelim: Tuesday, March 11, 2025 (7.30pm ET), Bailey Hall 101
Final: TBD
Grading
Final grades are based on homework assignments, programming projects, and the exams. For the 5780 level version
of the course, the research comprehension quizzes will also factor in.
For CS 3780 your final grade consists of:
Exams: 48%
Homework: 15%
Projects: 37%
For CS 5780 your final grade consists of:
Exams: 45%
Homework: 10%
Projects: 35%
Paper comprehension: 10%
Undergraduates enrolled in CS 3780 may choose to do the paper comprehension assignments; if completed, you will
receive the higher of your two grades between the above schemes.
Schedule
A tentative schedule is as follows, and includes the topics we will be covering, relevant reference material, and
assignment information. It is quite possible the specific topics covered on a given day will change slightly. This
is particularly true for the lectures in the latter part of the course, and this schedule will be updated as
needed.
NOTE: The due dates listed are generally accurate but may be subject to change. Tentative dates are
indicated
using unfilled boxes. While the release and due dates are
typically marked on
Tuesdays or Thursdays,
the actual dates might differ (e.g., a homework release marked for Thursday may instead be released on Friday,
with the due date
adjusted accordingly). Think of these as approximate timeframes for when assignments will be released or due.
Make sure to check Canvas for the exact assignment due dates.
While this course does not explicitly follow a specific textbook, there are several that are very useful
references to supplement the course.
Books
We will not be explicitly following any single textbook in this course. Nevertheless, the book by Murphy
roughly covers the material for the course and is recommended. Most suggested readings are assigned out of
these two texts.
[PML] Probabilistic Machine Learning: An Introduction, by Murphy
We will provide section numbers to this text alongside many of the lectures. This text is available digitally
through the Cornell University Library and a draft version is available directly from the author.
PML book website
[ESL] The Elements of Statistical Learning by Hastie, Tibshirani, and Friedman
This text provides a comprehensive introduction to statistical learning and provides in-depth discussion of
many of the topics in this course. The book is available directly from the authors.
ESL book website
Additional references
Three additional texts are provided that complement these texts and are useful for further study (or to
gain another perspective).
An Introduction to Statistical Learning by James, Witten, Hastie, and Tibshirani
(book website, also accessible via the Cornell library)
Patterns, Predictions, and Actions by Hardt and Recht (book website)
Fairness and Machine Learning by Barocas, Hardt, and Narayanan (book
website)
Background references
Background in linear algebra, probability, and calculus, as well as some "mathematical maturity"
is assumed for this course. If you feel you need a refresher, or would like to learn more about these topics, the
following resources may be useful.
A linear algebra review and reference from Stanford CS 229 (online)
Applied Machine Learning course materials
(taught at Cornell Tech by Volodymyr Kuleshov; online)
For recent work (on theory and applications) in machine learning, look at publications in
NeurIPS,
ICML, and
ICLR.
Additionally, arXiv is a good resource to keep up with the latest research
(note: articles may not be peer-reviewed).
FAQ
Q. I am currently on the waitlist [at position XX], do I need to do anything (e.g., email the instructors)?enrollment
A. No, you do not need to contact the instructors or the course staff. Given the class capacity, we
expect most people
to make it off the waitlist. However, even with space, you'll have to wait for CIS admins to add you to the
course (which is usually done
in batches).
For more information (the FAQ on position in the waitlist might be of
specific interest),
the following links:
can assist you. If you have any more questions, please email courses@cis.cornell.edu.
That said, even if you are not enrolled, we strongly recommend you attend the first lectures (to note the
specifics of the placement
exam), so long as there are physical seats available.
Q. I am having trouble joining the waitlist, what should I do?enrollment
A. If you are having trouble joining the waitlist, please double check that you have carefully followed all
the instructions noted in
If you are still having trouble, please email courses@cis.cornell.edu.
Unfortunately, the course
instructors or staff won't be able to help you with this.
As noted in the previous question, even if you are not enrolled, we strongly recommend you attend the first
lectures—we will detail the
specifics of the placement exam (and how you can access it, irrespective of your enrollment status) in the class.
Q. Prerequisite or corequisite: I'm currently enrolled in a course listed as a prerequisite, am I still allowed
to enroll?enrollment
A. We strongly recommend that you have completed the prerequisite courses prior to taking CS 3780.
However, if you are currently enrolled in a prerequisite course, we will use the placement exam scheduled during the
first week of class
to assess your comfort level with the necessary background knowledge for CS 3780.
Q. Can I take the class S/U?enrollment
A. We allow for S/U grading. From our perspective, all students are treated as though they are taking the
class for a letter grade;
we will then convert the letter grade to an "S/U" grade at the end of the semester using the conversion:
C- or higher = S, D+ or lower = U.
(Please be sure to review the collaboration policy noted below.)
Q. What about auditing the class?enrollment
A. We do not officially (via enrollment on the Student Center) allow for auditing the class. That
said, just sitting in on lectures
is fine, so long as there is physical space in the room. To conserve resources, we will not be able to provide
access to the course materials
(e.g., Canvas, Ed, etc.).
Rationale: We strongly believe that active, hands-on learning (from solving homeworks/assignments, in addition to
listening to lectures) would
be invaluable in gaining both in-depth and practical understanding of machine learning concepts.
Q. Are there other machine learning courses offered this semester?enrollment
A. See https://machinelearning.cis.cornell.edu/pages/courses.php
for machine learning courses offered at Cornell (please note that the list may be outdated and/or some courses may
not be offered this semester).
Here are a few CS 3780 equivalent courses that cover roughly the same core topics:
STSCI 3740/5740: Data Mining and Machine Learning (previously STSCI 4740; roster entry)
ORIE 3741/5741: Learning with Big Messy Data (previously ORIE 4741; roster entry)
Below is a self-compiled, non-exhaustive list of other possible courses covering machine learning topics (broadly
speaking) that are offered this semester:
CS 3780 is not a (strict) pre/corequisite:
INFO/CS 4300: Language and Information (roster entry)
CS 4670/5670: Intro to Computer Vision (roster entry)
CS 4744: Computational Linguistics I (roster entry)
CS 4740/5740: Natural Language Processing (CS 3780 is not listed as a pre/corequisite, but please check with
the instructors; roster entry)
CS 4701: Practicum in AI (satisfies project component, CS 3780 is not a strict prerequisite; roster entry)
CS 3780 or equivalent is a prerequisite (recommended after CS 3780):
CS 4782/5782: Intro to Deep Learning (roster entry)
CS 4789/5789: Intro to Reinforcement Learning (roster entry)
Q. What happened to CS 4780? Or, why is it CS 3780 now?logistics
A. Nothing, really. We renamed CS 4780 to CS 3780 as part of the introduction of the AI Minor. The goal was
to signal that this is one
of the classes to take first as your entry into Machine Learning (ML), and that other classes can build on CS
4780—ah, we mean CS 3780—as a prerequisite.
The content of the class has not changed in a substantial way, except that we have made sure that ECE 3200
(previously ECE 4200), ORIE 3741 (previously
ORIE 4741), STSCI 3740 (previously STSCI 4740) now all cover the same core content as CS 3780. So, these classes are
now largely interchangeable, and
you can take any one of them.
(Q/A adapted from the Fall 2024 run of the course—the first
offering of the course as CS 3780.)
Course policies
Inclusiveness
You should expect and demand to be treated by your classmates and the course staff with respect. You belong here,
and we are here to help you learn and enjoy this course. If any incident occurs that challenges this commitment to
a
supportive and inclusive environment, please let the instructors know so that the issue can be addressed. We are
personally committed to this, and subscribe to the
Computer Science Department's Values of Inclusion.
(This statement was reproduced with permission from Dan Grossman.)
Mental health resources
Cornell University provides a comprehensive set of mental health
resources and the student group Body Positive
Cornell has put together a flyer (for 2022-23; resources mentioned
are still relevant) outlining the resources available.
Class participation
You are encouraged to actively participate in class. This can take the form of asking questions in class,
responding to questions to the class, and actively asking/answering questions on Ed, etc.
Collaboration policy
Students are free to share code and ideas within their stated project/homework group for a given assignment, but
should not discuss details about an assignment with individuals outside their group. If you are taking
the course
for S/U grading, please be sure to manage your expectations with your assignment partners about your
intended level of
commitment.
The prelim and final are individual assignments and must be completed individually (and not as a group).
Academic integrity
(In the context below, "you," refers to yourself when work is to be done individually, or to your group when work
is to
be done in a group. The statement below is reproduced with explicit permission from David Bindel.)
An assignment is an academic document, like a journal article. When you turn it in, you are claiming everything in
it is your original work, unless you cite a source for it.
If you get an idea from a classmate, the instructors, a book or other published source, or elsewhere, please
provide
an appropriate citation. This is not only critical to maintaining academic integrity, but it is also an important
way
for you to give credit to those who have helped you out. When in doubt, cite!! Code or write-ups with appropriate
citations will never be considered a violation of academic integrity in this class (though you will not receive
credit
for code or write-ups that were shared when you should have done them yourself).
For more information, please refer to the Cornell's Code of Academic
Integrity.
Note: This course is participating in Accepting Responsibility (AR), which is a pilot supplement to the
Cornell
Code of
Academic Integrity (AI). For details about the AR process and how it works with the AI Code, see the
AR website.
Use of generative AI tools
Given the ubiquity of generative AI tools (such as OpenAI ChatGPT, GitHub Copilot, Meta Llama, Google Gemini,
etc.), and this being a
machine learning course, we permit the use of such tools only under the following conditions:
Given that this is an "intro" (to machine learning) class, we require that the first draft of the solutions be
your own—you can
refine (not generate) your solutions using generative AI tools.
If you use generative AI tools, you must (a) explicitly cite the tools used and (b) provide the exact prompts
used in refining
your solution.
It will be considered a violation of the academic integrity to use generative AI tools to generate initial drafts,
or to refine solutions
without providing the necessary citations. That said, if you choose to use generative AI tools, we recommend SAI
(use this link to join CS
3780 SAI),
as it allows us, instructors, to upload course materials to SAI for more relevant generations.
Grading and regrade requests
There is no preset forced distribution of grades in this course—we would be thrilled if everyone scored an A+, and
we believe
you're all capable of doing so.
For each homework or exam that is graded, we will make preliminary score-to-letter-grade landmarks (NOT
cutoffs) available to you.
These landmarks are homework- or exam-specific and are determined based on the difficulty of the homework or exam.
Students are not in competition with each other, i.e., your grades are not determined by how well others do. As
such,
we will not report means, medians, or other statistics about the class performance. As Lillian Lee once said,
"[r]eporting
median is guaranteed to make at least half the class feel bad," even if everyone did well.
Regrade requests: Given the large class size, only those regrade requests made via Gradescope will be
considered; this
facilitates a centralized record of all regrade requests. We reserve the right to disregard regrade requests made
via other channels
(e.g., Ed, email, office hours, etc.).
We want grades to accurately represent your understanding of the course material. If you believe that a grading
error was made,
please submit a regrade request on Gradescope. Be sure to explain clearly why you believe your answer deserves
more credit.
That said, we explicitly note that if we notice that we awarded you more points than you should have been, we are
honor-bound to
rectify such scores to the correct values.
Electronic device policy
Use of electronic devices such as laptops and tablets will not be permitted during class (with the
exception of specific
in-class activities and for note-taking purposes). We are not plain evil, we are just following extensive research
on the negative
effects of in-class laptop use on learning. That said, for your convenience, printed notes (i.e., physical copies)
will be made
available at the start of the lecture for you to follow along.
(The statement above was adapted from INFO/CS 4300 course syllabus, with the permission of Cristian
Danescu-Niculescu-Mizil.)
SDS accommodations
In compliance with the Cornell University policy and equal access laws, we are available to discuss appropriate
academic accommodations that may be required for student with disabilities. Requests for academic accommodations
are to be made during the first three weeks of the semester, except for unusual circumstances, so arrangements can
be made. Students are encouraged to register with Student Disability Services to verify their eligibility for
appropriate accommodations.
If you need immediate accommodations, please speak to us, instructors, after the class or send an email to us:
intro-ml-prof@cornell.edu and SDS:
sds_cu@cornell.edu.
Version history (Or, what changed?)
Any major changes (i.e., changes that require your attention) made to the website are logged below with the
corresponding dates.
