Probability and Mathematical Statistics in Data Science
Spring 2022 - Shobhana Stoyanov
Probability and Mathematical Statistics in Data Science
Spring 2022 - Shobhana Stoyanov
All aspects of the course are subject to change, because of the changing scenarios this semester.
In this course, we will state precisely and prove results discovered while exploring data in Data 8. Topics include: probability, conditioning, and independence; random variables; distributions and joint distributions; expectation, variance, tail bounds; Central Limit Theorem; symmetries in random permutations; prior and posterior distributions; probabilistic models; bias-variance tradeoff; testing hypotheses; correlation and the regression model.
You must enroll in the lecture and in a discussion section that meets twice weekly. If your section time is incompatible with your time zone, you are allowed to attend other discussion sections without changing which one you are registered for. However, we strongly encourage you to stick with one section or one uGSI for a more consistent learning experience. Some of the section times will also be rescheduled to accommodate students in different time zones. More information will be posted on Piazza.
The course textbook will be available on the course website. Lectures and the textbook will be closely related in content and sequence, though examples done in lecture might be different from those in the text. Solutions to the practice problems can be found here.
The current campus plan is for remote instruction until January 28. Accordingly, for the first two weeks, lecture and sections will be via Zoom. Note that these sessions will be synchronous, so lecture will be on Tuesdays and Thursday from 6:30 - 8:00 PM. All the zoom links can be found on the Piazza site for the course (more information below). The lectures will be recorded and posted for students who are unable to join the synchronous lecture. Note that there may be a delay in posting the lectures. GSI-led live discussion sessions will meet on Mondays and Wednesdays at various times. While we will not take attendance during lecture and section, we highly encourage you to attend because they are amazing opportunities for you to develop problem-solving skills and ask clarification questions.
There will be a Piazza site for this class. We will post important announcements on Piazza so please check it regularly. We will not be using bcourses for this course, except to store the lectures in the Media Gallery for the course. Make sure to read the Piazza etiquette post!
(All dates and times are Pacific Time, so if you are not in Berkeley, it is your responsibility to convert them to your time and complete the submissions by their deadlines.)
We take Academic Misconduct very seriously. Anyone caught cheating on a weekly quiz or a homework will receive a 0 on that assessment, and be turned over to the Office of Student Conduct. The same goes for the exam. If you cheat on the final, you will receive an F in the course. Note that unauthorized redistribution of quiz or exam material to anyone or any online platform, during or even after the designated testing period, is not only academic dishonesty but also intellectual property infringement, and can result in suspension from the university.
Course grades will be assigned using the following weights:
Historically this class has a typical statistics course curve with 30% of the class getting some kind of A and 30% of the class getting some kind of B and 30% some kind of C. See here for the grade distributions in the past semesters. My curve is usually more generous than this, with about 65%-70% of the class getting at least a B-.
Late submission of homework will not be accepted under any circumstances, unless you have relevant university accommodations. If you have such accommodations, please provide the formal notification to your GSI before the end of the second week of classes. There will be no alternate due dates for assignments missed due to illness, other commitments, and so on. The drops are intended to cover those situations.
You are encouraged to discuss practice problems and homework with your fellow students and with course staff. Arguing with friends about exercises is an excellent and time-honored way to learn. However, you must write up all your assignments on your own.
Copying assignments from others is not only dishonest, it also doesn’t help anyone. Each exercise requires its own combination of ideas, and each student needs practice in coming up with those combinations, or else they will be at a loss when trying to use probability theory in their future work. From a purely practical perspective, all students must work independently on Stat 88 quizzes and exams – no collaboration allowed. If a test is the first time a student works independently, then the test is not likely to go well.
Stat 88 materials including exams and solutions are the intellectual property of the course developers. From the campus statement on Academic Integrity: “… students may not circulate or post materials (handouts, exams, syllabi, i.e. any class materials) from their classes without the written permission of the instructor.” I am extremely tough with dishonest students and I hope that I will not be put in that situation in Stat 88. I expect that you will work with integrity and with respect for other members of the class, just as the course staff will work with integrity and with respect for you.
If you have test (or other) accommodations, please make sure that I have received an email about it from your DSP advisor. It is your responsibility to make arrangements in a timely manner through DSP so that we can make the appropriate accommodations.
If you have any comments or suggestions, please feel free to send an email to me or your GSI. All feedback is welcome.