Stat 102B-Intro Computation and Optimization for Statistics/Sanchez

Course Syllabus

(a) Professor: Dr. Juana Sanchez, Ph.D., [email protected], http://directory.stat.ucla.edu/faculty/single-page/?smid=159 When e-mailing, please indicate the class (Stat 100A) and also your full name and ID.

(b) Lectures: Lectures for this class are synchronous most of the time and recorded live. Sometimes the lectures will be asynchronous (the latter meaning that the videos of the lectures for that lecture day are pre-recorded and lecture time will be used for Q&A about the content of the videos and practice. Two additional office hours of the Professor at times different than the lecture times will be offered each week. Lecture slides templates will be posted before each lecture, and students will complete them as they watch the video. All lecture materials are required.

(c) TA session: The TA session will be synchronous, i.e., you will meet with the TA in real time via zoom. Attendance is not required but it is highly recommended. You will complete a posted handout before the TA session, and given a chance to discuss it with the TA during session. That handout will be submitted.

(d) Professor’s Office hours: There will be two hours of office hours each week. They will be posted in CCLE. There will also be e-mail office hours: daily, M-F, 8 AM to 6:00 PM excluding Saturday and Sunday. The fastest way to obtain help is contacting Dr. Sanchez. We understand that each student is unique and needs different kind of help with the course material.You are encouraged to use also the discussion forum.

(e) Textbook and required readings There will be a series of required readings from the library that you may access for free if you install VPN to access the library. You may get help setting VPN from BOL Help Desk. In regular times it is located at Kerckhoff Hall, but you may call them at 310-267-4357, or go to www.bol.ucla.edu

(f) Pre-requisites: Stats 100B and Stats 102A and Math 33A.

(g) Enrollment requirement: You must be enrolled in Stat 100A to access the web site, take exams and have work count towards grade. Work missed due to not being enrolled will not be made up.

(h) Assessment, milestones and deadline:

● quizzes (15%): The lowest quiz scores will be dropped. These are quizzes to do each week and will be due at the end of the week. Due to the nature of these quizzes, which is to assess your understanding of the lectures and readings in the book, the activities done during the TA session, and get feedback on progress, they are a very important component of the class. Quizzes can not be made up. If you do not do them at all you lose the 15%. Even if you decide to not do them and lose the points, you should click on them when they are due, open them and submit with nothing to be able to access them to review for exams. We will not create duplicates of assessments just for you because you decided not to open the quizzes when they were due.

● Homework (25%). Lowest score dropped. Required. No make ups. Homework consists of problems from the textbook and/or based on the material discussed in lecture, but it may also consist of watching a video and writing about it, or reading some material, or other formats. Due dates for all homework is Friday before 11 PM. Homework can not be made up, that is, if you do not do it you lose the 25%. Even if you decide to not do them and lose the points, you should click on them when they are due, open them and submit with nothing to be able to access them to review for exams. We will not create duplicates of assessments just for you because you decided not to open the homework or other assessments when they were due.

● Midterm (25%) Required. Comprehensive. No make ups. No late allowed. The midterm time window within which you can start the exam will open on Friday, August 27th at 8 AM and will close on Saturday, August 28th at 8 AM, Los Angeles time. More information and detailed instructions will be posted at the time of the exam. Not taking the midterm exam results in an F in the course. Some practice midterms are posted in the Cognella active learning site.

● Final exam (30%) Required, no make ups, comprehensive. The final exam time window starts on Friday, Septem-ber 17th, at 8 AM and ends Saturday, September 18th, at 8 AM. The exam is a two hour exam but you will be given three hours within that time window. No make ups, no late allowed. Not taking the final exam results in an F in the course. Detailed instructions will be posted at the time of the exam.

● Discussion forum: (5%) Our course web site in CCLE comes with a discussion forum. We may use this and other forum such as slack to keep track of what we are doing during the quarter, working together and keeping up to date with the course material. Students helping each other by explaining correctly a concept that others ask about, or directing students to the lecture notes, or textbook pages where the concept is explained, will get some extra credit if the response is correct. The forum is not substitute for you to ask Dr. Sanchez directly. And, please, for other matters unrelated to course material’s content, email Dr. Sanchez directly [email protected], http://directory.stat.ucla.edu/faculty/single-page/?smid=159

You should know that the fastest way to get an answer from Dr. Sanchez about anything regarding the course or homework, quizzes or class content is by email. E-mail office hours are 8AM to 6 PM daily except Saturday and Sunday and excluding teaching times.

(i) Contact Dr. Sanchez only (not the TA) the first week of class, if you have any questions about the course requirements, deadlines or any other administrative matter. Email Dr. Sanchez. [email protected], http://directory.stat.ucla.edu/faculty/single-page/?smid=159

(j) Questions about grading in an assignment or exams must be brought to the attention of the professor (not the TA) no later than two days after the exam or homework has been returned in class or grade posted by the professor (not after you look at it another day). Make an appointment by email sending as many dates and times as possible. Office hours are not used for talking about grading or your standing in the class. [email protected], http://directory.stat. ucla.edu/faculty/single-page/?smid=159

(k) Your grade will be determined as follows: A+ 98-100; A 95-97; A- 90-94; B+ 88-89; B 82-87; B- 80-81 C+ 78-79; C 72-77; C- 70-71; D+ 68-69; D 62-67; D- 60-61; F Below 60; P (C or better); S (B or better) The scale will be adjusted at the discretion of the professor.

(l) Reading the whole syllabus is required. It answers many of your administrative questions.

(m) Academic integrity and class policy

● Conduct. As a student and member of the University community, you are here to get an education and are, therefore, expected to demonstrate integrity in your academic endeavors. All students must uphold Univer-sity of California Standards of Student Conduct as administered by the Office of the Dean of Students http://www.deanofstudents.ucla.edu/individual-student-code Students are subject to disciplinary action for several types of misconduct or attempted misconduct, including but not limited to dishonesty such as cheating, mul-tiple submission, plagiarism, or knowingly furnishing false information.

● Students needing academic accommodations based on a disability must contact immediately the Center for Accessible Education (CAE) at (310)825-1501 or present in person at Murphy Hall A255 at the beginning of first week of classes. As the professionals delegated authority from the campus to determine reasonable disability accommodations, CAE will assess all requested accommodations and communicate appropriately with faculty. After you have contacted CAE, they will contact the professor. But you should also let the professor know that CAE will be in touch. For more information visit www.cae.ucla.edu

● Community expectations. UCLA’s office for Equity, Diversity, and Inclusion provides resources, events, and information about currect initiatives at UCLA to support equality for all members of the UCLA community. I hope that you will communicate with me or your TA if you experience anything in this course that does not support an inclusive environment, and you can also report any incidents you may witness or experience on campus to the Office of Equity, Diversity, and Inclusion on their website https://equity.ucla.edu

Learning goal of this course

The goal of this course is twofold: first, to make you appreciate the role of vector spaces, matrix algebra and calculus in optimization methods and their numerical counterparts as applied in data science and the statistical foundation of data science that you are learning in the Stats Department (such as statistical inference and regression methods); second, to realize that learning the bare minimum that you will learn in this class will help you understand AI, ML and algorithmic decision making concepts in more depth. That way you will be able to easily switch technology stacks, frameworks and programming languages when changing jobs in the future.

Learning outcomes of this course

The professor reserves the right to modify the order and/or content of this outline

By the end of this course, students will be able to do each of the following:

(I) Introductory level operations with vectors, matrices and calculus relevant for Statistics and Data Science, and their implementation and storage in R.

(a) Vector operations and how data science metrics are written in vector form.

(b) Random vectors and their expectations vectors for stochastics.

(c) Matrix operations and how typical summary statistics are written in matrix-vector form. Variance-covariance matrix, Hessian matrices, gradients for optimization.

(d) Algorithms to invert matrices, in particular variance-covariance matrices for multivariate normal distribu-tions and other distributions.

(e) Single and Multivariate Optimization theory and algorithms to implement it.

(II) Use the operations learned with vectors, matrices and calculus and optimization theory to common machine learning, artificial intelligence and deep learning problems. Examples are:

(a) calculating clusters from the matrix operations (unsupervised learning)

(b) Playing with distance matrices for hierarchical clustering algorithms

(c) K-means clustering algorithms

(d) Model based clustering with the EM algorithm.

(e) Least squares minimization for regression-based methods.

(f) Single and multiple parameters maximum likelihood estimation and least squares

(g) Artificial Intelligence optimization applications.

(III) Use the operations learned with vectors, matrices and calculus and optimization theory for dimension reduction

(a) Principal component analysis and other dimension reduction methods.

(IV) Classification (supervised learning)

(a) Classification methods.

Required software

        Required software, Rstudio: RStudio installed in your computer or access it from the laptops that you can check out in the library. CCLE web site will have information on that.https://www.rstudio.com/

        The comprehensive R archive https://cran.r-project.org/R for documention is also required.

Some tips about learning, group work, reading and thinking ahead and study habits

        In addition to the special circumstances advice that you have received for learning online from https://www. teaching.ucla.edu/resources/student-remote-learning you should consider the subject-matter-specific ad-vice given below:

● Our aim with all lecture-time activities, lectures, other required reading, practice homework, quizzes, exams and TA sessions is to acquire enough breadth and depth of the subject to incrementally gain deeper understanding and mastering of the subject and become more intelligent using it. All students with the prereqs can achieve that. Communicating and speaking about the subject matter, reading and preparing ahead of time to asses what you can do on your own, is an important component to achieve our learning goals.

● Talking about what we learn, in our own language, has been found to improve learning and retention. That is why you are required at times to work with students in the class and to participate in class. Learning takes place in society,we are not isolated. Talking about what we learn enhances retention and comprehension. This particularly helps become more familiar with the jargon of probability. Writing your thoughts and methods has the same effect. Asking questions also.

● Your active engagement in your learning, organized studying, reviewing the notes you take in lecture, the lecture notes and reading required, connecting past and new material your own way, as well as self-assessment and good study habits have been found to be very important factors in students‘ success in their academic endeavors. The homework, quizzes, exams and activities in class and the TA sessions, are intended to help all of you become gradually feel more and more comfortable thinking and speaking like those in the field.

● Research has found that students who immediately or as soon as they can after a class, read the notes they took, organize and write in their own way what they learned that day, have questions and seek answers about what they do not understand, and keep up to date with the course web site, think about what is and is not clear, and solve their doubts sooner by asking, are more successful learner. Talk to each other about your study habits. Review, record, reflect, report.

● By working in groups and talking to each other about the material being learned all win. Working on a problem together, preparing a concept map, trying to understand conflicting arguments, and pros and cons of actions, helps retention of the material learned.

● When asked to work alone, you are expected to do so.