CIS 5150 Math For Machine Learning

Spring 2023

Instructors
Jean Gallier, Professor, Department of Computer and Information Science
Jocelyn Quaintance, Lecturer, Department of Computer and Information Science

Course Description
There are hardly any machine learning problems whose solutions do not make use of linear
algebra. This course presents tools from linear algebra and basic optimization that are used to
solve various machine learning and computer science problems. It places emphasis on linear
regression, data compression, support vector machines and more, all of which will provide a
basis for further study in machine learning, computer vision and data science. Both theoretical
and algorithmic aspects will be discussed, and students will apply theory to real-world situations
through MATLAB projects.

Course Learning Objectives

● Understand the important mathematical concepts of linear algebra and how they
underpin real-world problems.
● Apply MATLAB to address curve drawing, data compression and optimization problems.
● Explain the mathematical foundations which guarantee the existence of a singular value
decomposition (SVD) for a linear map, and be able to use several applications of the
SVD.
● Develop a conceptual understanding of the Lagrangian framework and apply it to
quadratic optimization problems found in machine learning.

Course Prerequisites
● Prerequisite: Calculus Chapters 8, 9, 10, and 48 of Schaum’s Outlines of
Calculus fifth edition by Frank Ayers and Elliott Mendelssohn
● Suggested prerequisite: Undergraduate course in linear algebra (helpful but not
required)
● Suggested prerequisite: Chapters 1 through 3 of Schaums Outline of Linear
Algebra, fourth version by Seymour Lipschitz and Marc Lipson

CIS 5150 – Math for Machine Learning | Spring 2023 v1 | Property of Penn Engineering
Course Textbook
Recommended
● Linear Algebra and Optimization with Applications to Machine Learning Vols. I & II,
Gallier & Quaintance (Link to PDF will be provided)

Grading & Assessment

You must attempt all graded assignments to pass the course. If you have any questions
or concerns about grading or progress in the course, please reach out to the instructor.

This course will use a variety of assessments to determine whether learners understand
and can apply the key concepts and skills that the course teaches. This includes:

Type % Description

50% Projects #1 and #2 have A and B sections.

Projects Projects #3, #4, and #5 do not. Students will work
in teams of size 2 or 3, and each team will submit
a joint project. All members of the team will
receive the same grade for a completed project.

Homeworks 35% There are 14 homeworks. All homeworks are

optionally group assignments (groups of size up to
3). We won’t be using the group Canvas tool
(students can work together but are graded
individually). Students will submit the same
assignment separately and TAs will grade them
the same. If you change team members, you
must let the TA's know about the change no
later than the day of submission.

Final Exam 15% The final exam will be cumulative and in week 15
of the course.

Each graded assessment will use Canvas/Gradescope for submission.

Please read the instructions for each assignment very carefully to make sure you
know what to submit to receive credit!

Late Policy/Extensions

CIS 5150 – Math for Machine Learning | Spring 2023 v1 | Property of Penn Engineering
The instruction staff is committed to your success and understands how challenging it can be to
learn online while balancing other commitments. Despite students’ best intentions, sometimes
life gets in the way and a little extra time to complete an assignment may be necessary.

If you need extra time on an assignment, you can obtain an extension for extenuating
circumstances. If an extension is not approved, an assignment that is turned in late will receive
a 15% grade reduction per day up to 3 days. After the 3rd day, the highest grade you can earn
will be 0%.

For extensions please fill out the extension request form found in the “Resources”
section in Canvas). Please fill out this form for any and all extension requests you wish
to submit. These extension requests must be submitted at least 24 hours before the
assignment deadline. If your request is granted, you will see updated deadlines
reflected in Canvas by Mondays at 5 pm ET. You will need to fill out this form for each
assignment you wish to have an extension for.

Regrade Requests
Regrade requests must be made through Gradescope. Regrade requests may take up to a
week to process at the discretion of the instructor..When submitting a regrade request, please
explain (in detail) why you feel the grading is incorrect.

Extra Credit

There is no opportunity for extra credit in this course.

Other Course Activities

The following activities are not mandatory, but will greatly support your success on the
graded assignments.

Recitation
Recitations will take place once every two weeks, and are live sessions with your TAs
that explain how the course content is being applied in the real world.If you are not able
to attend the scheduled recitation times, you can review a recording of the session,
posted the day after.

Additional Segments
The professor may add additional optional segments to support the class as needed.

CIS 5150 – Math for Machine Learning | Spring 2023 v1 | Property of Penn Engineering
Creating an Inclusive Environment
All members of the course community – the instructor, TAs, and students – are expected
to work together to create a supportive, inclusive environment that welcomes all
students, regardless of their race, ethnicity, gender identity, sexuality, religious beliefs,
physical or mental health status, or socioeconomic status. Diversity, inclusion, and
belonging are all core values of the MCIT Online program, the instruction staff, and this
course. All participants in this course deserve to and should expect to be treated
with respect by other members of the community.

Discussion boards, messaging channels, recitations, office hours, and group working
time should be spaces where everyone feels welcome and safe. In order to facilitate a
welcoming environment, students of this course are expected to:
● Exercise consideration and respect in their speech and actions
● Attempt collaboration and consideration, including listening to opposing
perspectives and authentically and respectfully raising concerns, before conflict
● Refrain from demeaning, discriminatory, or harassing behavior and speech

All members of the course community are expected to be familiar with and abide by the
University’s guidelines on general conduct and sexual harassment:
● University Code of Conduct:
https://catalog.upenn.edu/pennbook/code-of-student-conduct/
● University Sexual Harassment Policy:
http://www.upenn.edu/affirm-action/introsh.html

Students should also be familiar with other University guidelines regarding personal
conduct:
● Conduct & Personal Responsibility guidelines in Pennbook:
https://catalog.upenn.edu/pennbook/#policiesbytopictext
● University Principles of Responsible Conduct:
http://www.upenn.edu/audit/oacp_principles.htm

If you are a victim of, witness, or are otherwise affected by unacceptable behavior:
● In cases of sexual harassment or assault, please consult DPS Special Services
(https://www.publicsafety.upenn.edu/about/special-services/sensitive-crimes/ at
215-573-3333; this is a confidential resource.
● To report other bias incidents, contact the Penn Office of Diversity:
https://diversity.upenn.edu/diversity-at-penn/bias-motivated-incident-report
● For other violations of the code of student conduct, the Office of Student Conduct
has an incident reporting form at https://www.osc.upenn.edu/referral-form

CIS 5150 – Math for Machine Learning | Spring 2023 v1 | Property of Penn Engineering
If you are unsure which office to contact, please contact the instructor or any Penn
Engineering Online Learning staff member.

Getting and Giving Help

TA and Faculty Support
TAs will hold weekly private live drop-in office hours. Live drop-in private office hours are
indicated on the course calendar and held through OHQ.io. TAs will help a queue of
students on a first-come-first served basis and meet one on one with students.

Your professors will hold regular Open Office Hours every week, and maybe available
for a limited number of private meetings per week, depending on the needs of the class.

Collaboration Guidelines
In the professional world of software development, collaboration—including using code
that others have written—is both practical and ubiquitous.

However, to prepare for entering that professional context, you need to develop a full
set of software development skills so that you are both able to create your own code
and evaluate the quality of someone else’s code that you might use.
In the context of this course, independent work and evaluation are critical. Do not
collaborate with others on individual graded assignments unless it is explicitly
indicated. The inappropriate collaboration will be considered cheating and considered
under Penn’s Code of Academic Integrity.

Discussion forums and recitations are collaborative—please take advantage of those

times to work with your colleagues. For general communication with your colleagues,
use your Slack channels or Slack direct messages.

Forming study groups to understand the material is also a good idea, as long as you
stay on the conceptual level and are not collaborating on the graded assignments
directly.

Note: When in doubt always ask the instructor or TA first, to avoid any potential
collaboration that can lead to academic dishonesty.

Do not cheat. Please note that searching for solutions online is the same as
cheating. Posting solutions online is also considered cheating. If you are caught

CIS 5150 – Math for Machine Learning | Spring 2023 v1 | Property of Penn Engineering
posting solutions or code to a publicly accessible location (like StackOverflow or
GitHub), it will be considered cheating. If you do use GitHub (or similar cloud-based
code management system) to set up a remote code repository, YOU ARE REQUIRED
TO KEEP THAT REPOSITORY PRIVATE.

You can further read Penn's Code of Academic Integrity page on this subject
matter, as well as the SEAS Graduate Student guidelines on the code of ethics.

Plagiarism Policy

The first instance of homework plagiarism will be handled by the instructor and may
include escalation to the Office of Student Conduct.

Second instances or exam plagiarism will be turned over immediately to the University
of Pennsylvania Office of Student Conduct.

Regardless of previous work in the course, the penalty for plagiarism is the failure of the
course (regardless of current course average), and potential permanent notation on
your academic record that will follow you to all future academic institutions and possibly
future employers. If you are unfamiliar with what constitutes plagiarism at Penn, visit
Penn’s Code of Academic Integrity. Please note that searching for solutions or code
online is a violation of academic integrity. Sharing solutions or code with another
student (unless working on a group project or other collaborative assignment) is also
a violation of academic integrity. This includes posting solutions and code publicly
online, even after you've completed the course. If you discover publicly viewable
solutions for the assignments of this course, please let the course staff know immediately.

Access to Materials and Content Before and After Graduation

If you would like to retain copies of your submitted assignments, you must download them
from Gradescope, Canvas, Codio, and any other platforms that you submit to during the
semester in which you are taking that course. Your SEAS account will be permanently
deleted in the fall following your graduation. Once your SEAS account is deleted, we will be
unable to re-grant you access to the MCIT Online Program materials and course content.
Therefore, if you would like to retain the content from each course after you graduate
from the program, you must download the content locally to your computer during
the semester that you are in the course.

CIS 5150 – Math for Machine Learning | Spring 2023 v1 | Property of Penn Engineering
Recording Notice

Public office hours, recitations, and other live events will be recorded, used, and may be
made available to class participants during the current semester as well as students
who take the class in future semesters.

Private office hours will also be offered and are not recorded. Students who do not wish
to attend the publicly-recorded office hour may attend the private office hours.

Spring 2023 Course Schedule and Important Dates

Dates are subject to change. Please check Ed Discussion for announcements regarding schedule
changes.
CIS 5150 | Spring 2023
Note: Weeks run Monday through Sunday
WEDNESDA
SUNDAY MONDAY TUESDAY Y THURSDAY FRIDAY SATURDAY
Jan 8 WEEK 1 > 9 10 11 12 13 14
First Day of
Term

15 WEEK 2 > 16 17 18 19 20 21
Martin
Luther King,
Jr. Day HW1 Due

22 WEEK 3 > 23 24 25 26 27 28
HW2 Due

29 WEEK 4 > 30 31 February 1 2 3 4

Project 1A
Due
HW3 Due
5 WEEK 5 > 6 7 8 9 10 11
HW4 Due

12 WEEK 6 > 13 14 15 16 17 18

CIS 5150 – Math for Machine Learning | Spring 2023 v1 | Property of Penn Engineering
 Project 1B
Due
HW5 Due
10 WEEK 7 > 20 21 22 23 24 25
HW6 Due

26 WEEK 8 > 27 28 March 1 2 3 4

Project 2A
Due
HW7 Due
5 6 7 8 9 10 11
Spring Term Break (No Classes)

12 WEEK 9 > 13 14 15 16 17 18
Project 2B
Due
HW8 Due
WEEK 10 >
19 20 21 22 23 24 25
HW9 Due

WEEK 11 >
26 27 28 29 30 31 April 1
HW10 Due

2 WEEK 12 > 3 4 5 6 7 8
Project 3 Due
HW11 Due
WEEK 13 >
9 10 11 12 13 14 15
HW12 Due

WEEK 14 >
16 17 18 19 20 21 22
Project 4 Due
HW13 Due

CIS 5150 – Math for Machine Learning | Spring 2023 v1 | Property of Penn Engineering
 WEEK 15 >
23 24 25 26 27 28 29
HW14 Due Project 5 Due
Final exam ----------->
30 May 1 2 3 4 5 6
--------->

CIS 5150 – Math for Machine Learning | Spring 2023 v1 | Property of Penn Engineering