Syllabus
Table of contents
About This Course
- How do we know if an avocado is going to be ripe before we eat it?
- How do we predict the salary of a future data scientist?
- How do we teach a computer to read handwriting?
The world is increasingly recognizing the value of data in solving complex and open-ended problems such as these. Instead of explicitly telling the computer exactly how to differentiate between the letters of the alphabet, we instead give the computer many examples of each letter and let it learn the differences automatically. Similarly, by identifying patterns in data, we can learn which factors combine to make an avocado ready-to-eat or a person likely to be a successful data scientist. The explosive growth of data science is largely due to the fact that this approach of learning from data often works remarkably well.
But how do we learn from data? This is the central question of DSC 40A. We will see that virtually every rigorous learning method involves two steps: 1) turning the abstract problem of learning into a concrete math problem; and 2) solving that math problem. This quarter, we will see how to apply this fundamental approach in a variety of contexts. After this class, you will understand the basic theoretical principles underlying almost every machine learning and data science method — from simple linear regression to deep neural networks. You’ll also be better prepared to tackle the math you’ll see in your upper-division courses, like vector calculus, linear algebra, and probability.
Communication
This quarter, we’ll be using Campuswire as our course message board. You should be added to Campuswire automatically; if not, you can add yourself. Please join right away as we’ll be making all course announcements through Campuswire.
If you have a question about anything to do with the course — if you’re stuck on a problem, didn’t understand something from lecture, want clarification on course logistics, or just have a general question about data science — you can make a post on Campuswire. We only ask that if your question includes some or all of an answer (even if you’re not sure it’s right), please make your post private so that others cannot see it.
Course staff will regularly check Campuswire and try to answer any questions that you have. You’re also encouraged to answer questions asked by other students. Explaining something is a great way to solidify your understanding of it!
We won’t be using the direct messaging (DM) functionality of Campuswire, nor will we use email to answer questions about the course. Please don’t DM or email staff members, just make a private or public Campuswire post instead!
Technology
We will be using several websites this quarter. Here’s what they’re all used for:
- Course Website: where all content will be posted.
- Campuswire: discussion forum for announcements and communication.
- Gradescope: platform for submitting assignments and viewing grades. You should be automatically added to Gradescope; let us know if not.
- Datahub: UCSD’s data science and machine learning platform, for coding in Jupyter notebooks.
Course Structure
This course will include lectures via zoom, groupwork sessions in discussion section (also via zoom), and weekly homework assignments.
Lecture
Lecture is meant to introduce you to the main concepts of the course. Attendance is highly encouraged and positively correlated with success in the course, though it will not be required. Attending lecture gives you the opportunity to ask questions, answer ungraded concept-check polls, and participate in discussion.
Lectures will be posted on campuswire within a few hours, we will provide links on the course webpage.
In-person attendence is not required unless a midterm or final is taking place.
See the homepage of this website and the resources tab for access to helpful resources that will help solidify your understanding of concepts covered in lecture. These include videos, slides, readings, and sometimes code. These will be your primary resources in this class, as there is no formal textbook.
Discussion
Discussion sections will be primarily used to facilitate problem-solving in small groups with peers. There is one discussion section on Fridays at noon.
We will provide a worksheet of problems, which you will complete in a group of two to four students. The groupwork should help reinforce concepts from the lecture and prepare you to do the upcoming homework assignment.
Attendance at discussion section is required for full credit on the groupwork, but if you cannot attend, you may complete the groupwork worksheet in a self-organized group of two to four students outside of the discussion section for 95 percent credit.
If you have specific people in your section that you want to work with, you may work together, otherwise, the TA will pair you with other students.
Submit your worksheet to Gradescope by 11:59pm on Friday night. Only one member of each group should submit the worksheet, and they should indicate the names of all group members on Gradescope. Worksheets won’t be graded on correctness, but rather on good-faith effort. Even if you don’t solve any of the problems, you should include some explanation of what you thought about and discussed, so that you can get credit for spending time on the assignment. In order to receive credit, you must work in a group of two to four students for at least 50 minutes. You may not do the groupwork alone.
Homeworks
This class will have weekly homework assignments, which will be due to Gradescope on Tuesdays at 11:59pm.
Homeworks should be written or typed up and turned in by each student individually.
You may talk to other students in the class about the problems and discuss solution strategies, but you should not share any written communication. You can tell someone how to do a homework problem, but you cannot show them how to do it. One way to tell if you are respecting this boundary is to ask yourself whether your collaboration could take place over the phone. Additionally, the content of your verbal communication should involve the problem-solving strategy and approach, and you should not directly compare answers with classmates.
Talking through homework problems with other students can be very valuable for many reasons:
- You will learn about someone else’s thought process and learn new ways of solving problems that you may not have thought of.
- You will get practice explaining your ideas, which is a useful life skill, and important for job interviews.
- You will get practice thinking critically about whether someone’s proposed solution actually works, and you will learn how to poke holes in shaky arguments.
As a result of this collaboration policy, students may have similar approaches to problems, but they should not have similarly presented solutions, such as word choice.
For each problem you submit, you should cite your sources by including a list of names of other students with whom you discussed the problem. Instructors do not need to be cited.
We also encourage you to come to instructor and staff office hours for help on homework questions. The homework assignments for this class are quite challenging and most students are not able to successfully complete the homework from attending lecture alone. Make sure to use the resources provided on the resources tab of the course website, actively participate in groupwork sessions, and plan to attend office hours at least once a week. Even if you don’t have specific questions, you will likely get a lot out of conversing about the material. See the calendar tab of the course website for the most up-to-date schedule of office hours.
You may post homework-related questions on Campuswire, though your questions (and answers) should be about approaches, not answers. If your question includes some or all of an answer (even if you’re not sure it’s right), you must make your post private so that others cannot see it. We are not able to tell you whether your answer is correct.
Weekly Schedule
To summarize, here’s what a typical week will look like in the course (there may be some deviations from this due to holidays and exams; the most up-to-date deadlines will be on the course homepage):
| Sunday | Monday | Tuesday | Wednesday | Thursday | Friday | Saturday |
|---|---|---|---|---|---|---|
| Lecture | Lecture | Discussion | ||||
| Homework due 11:59pm | Groupwork due 11:59pm |
Exams
There will be one midterm and one final exam:
- Midterm: Wednesday, July 19, in-person during lecture at 8am in RWAC 0115
- Final: Friday, August 4th, in person at 8am (location TBD)
The final exam for this course will consist of two parts, which will be graded separately: part one will cover the material of the midterm, and part two will cover the material following the midterm. If you do better on part one than the midterm, then your score on that part will replace your score on the midterm.
Exams should be taken completely alone, with no collaboration or communication with any other person. We may utilize randomization and multiple versions to ensure the integrity of exams.
Policies
Grading
Here’s how we will compute your grade.
| Component | Weight | Notes |
|---|---|---|
| Homework | 40% | drop lowest |
| Groupwork | 10% | drop lowest |
| Midterm | 25% | |
| Final | 25% |
Late Policy, Slip Days, and Drops
Each student has three slip days to use throughout the quarter. A slip day can be used to extend the deadline of a homework assignment by 24 hours. You can use at most one slip day on any single homework assignment. Slip days can only be used for homework assignments.
Slip days are applied automatically at the end of the quarter, and you don’t need to ask in order to use one. It’s your responsibility to keep track of how many you have left. If you run out of slip days and submit a homework late, it may still be graded so that you’ll see what questions you missed, but the grade will be changed to a zero at the end of the quarter. If you use more than three slip days, we will count the first three late assignments, and any late assignments after that will get zero scores.
Slip days are designed to be a transparent and predictable source of leniency in deadlines. You can use a slip day if you are too busy to complete a homework on its original due date. But slips days are also meant for things like the internet going down at 11:58pm just as you go to submit your homework.
If you have something going on in your life that is impeding your ability to do your classwork on time, please reach out to us as soon as possible so we can work something out.
In addition to providing you with slip days, we will drop your lowest homework and lowest groupwork. This gives you some additional flexibility for unforeseen circumstances.
Students on the waitlist or who join the class late are expected to keep up with the work and submit assignments by the deadlines.
The stated policies will be strictly enforced out of fairness for all students.
Regrade Requests
You can ask for a regrade on any assignment if you believe that the grader made a mistake. Remember that clarity is a part of your score — if you had the right idea but were unable to clearly communicate it, you may still not deserve full credit. We ask that you please submit your regrade requests within one week of the assignment grade being released; you can submit regrade requests directly on Gradescope.
Incomplete Grades
In the unfortunate circumstance that you become sick, suffer a loss, or otherwise experience a significant setback that is outside of your control, you may be eligible for an Incomplete grade, which allows you to complete the rest of the work at a later time. If you are experiencing challenges due to circumstances outside your control, please contact me ASAP and we can discuss the best course of action. Note that an Incomplete does not allow you to re-do work that has already been completed, only to do work that hasn’t been completed, so it’s best to reach out right away.
Academic Integrity
In this class, we expect that you will work hard, utilize allowed resources to master the course material, and act with integrity. Learning partially remotely presents new challenges for academic integrity, making it more important than ever to act honorably and make sure that the work you are submitting is reflective of your knowledge and abilities.
The UCSD Policy on Integrity of Scholarship and this syllabus list some of the standards by which you are expected to complete your academic work, but your good ethical judgment is also expected. Ignorance of the rules will not excuse you from any violations.
For this class, the following activities, among others, are considered cheating and are not allowed:
- Sharing written homework solutions with other students, or viewing written homework solutions from another student.
- Looking or asking for answers to homework problems in other texts or sources, including the internet and AI tools such as ChatGPT and GitHub CoPilot.
- Collaborating on exams, checking answers on exams, or communicating with any other person while taking an exam.
- Using unauthorized resources on homeworks or exams, including solutions from past iterations of this course, and AI tools such as ChatGPT and GitHub CoPilot.
The following activities are examples of things that are allowed in this class:
- Discussing homework problems with classmates and the instructional staff.
- Reading about concepts from lecture in outside texts, including the internet, without looking for answers to specific homework questions. If you accidentally find related material in another source, you must cite the source on your homework and write up your answer without consulting the source. To do otherwise is plagiarism.
Remember, Academic Integrity is about doing your part to act with honesty, trust, fairness, respect, responsibility and courage. If you are suspected of dishonest conduct, you will be reported to the Academic Integrity Office. Violations of the academic integrity policy will result in failing the course, and the Dean of your college may place you on academic probation or suspend or dismiss you from UCSD. Academic integrity violations are serious and the risk is not worth it!
A note on letter grades
The following is adapted from CSE 160 at the University of Washington.
Grading for this class is not curved in the sense that the average is set at (say) a B+ and half of the class must receive a grade lower than that. If everyone does well and shows mastery of the material, everyone can receive an A (this would be awesome!). If no one does well (this is unlikely), then everyone can receive a C.
Grading for this class is curved in the sense that we do not have a pre-defined mapping from homework and exam scores to a final GPA. There is no pre-determined score (e.g., 90% of all possible points) that earns an A or a B or a C or any other grade. To determine the final grade, we will ask questions like “Did this student master the material?”. With that said, grades will not be any stricter than the standard grading scale (where an A+ is a 97+, A is 93+, A- is 90+, etc). For instance, the threshold for an “A” will never be higher than 93%.
Try your best not to worry about grades, and we’ll reciprocate by being fair. We’re in this together 😎.
Support
Accommodations
From the Office for Students with Disabilities (OSD):
OSD works with students with documented disabilities to review documentation and determine reasonable accommodations. Disabilities can occur in these areas: psychological, psychiatric, learning, attention, chronic health, physical, vision, hearing, and acquired brain injuries, and may occur at any time during a student’s college career. We encourage you to contact the OSD as soon as you become aware of a condition that is disabling so that we can work with you.
If you already have accommodations via OSD, please make sure that we receive your Authorization for Accommodation (AFA) letter by the end of Week 3 so that we can make arrangements for accommodations. Share your AFA letter with the instructor and the Data Science OSD Liaison, who can be reached at dscstudent@ucsd.edu.
Acknowledgements
Thanks to other instructors of this course who have made various contributions, including but not limited to Janine Tiefenbruck, Justin Eldridge, Suraj Rampure, Yian Ma, and Gal Mishne. Thanks also to the many tutors and TAs who have supported this course since its inception!