STAT 548 Course information

CSE 547/STAT 548 Course information

Instructor:Marina Meila mmp at stat dot washington dot edu

Optional Textbooks "Mining of Massive Data Sets" by Jure Leskovec, Anand Rajaraman and Jeff Ullman. For each lecture, I will point out the chapters are relevant. The book contents will be supplemented with material in the form of lecture notes.

Lectures: Mondays, & Fridays 2-3:20 in MLR 301

Instructor's office hours: TBD on zoom ??

TA: Medha Agarwal

Course web page: http://www.stat.washington.edu/courses/548/sp25 (this page) will be used to post lecture notes prior to lecture, and homeworks. The resources page is useful for addional reading and software resources. All other materials will be posted on Canvas. Everything posted here that is directly relevant to the learning objectives will be linked to from Canvas, hence you won't need to visit the web site once you are registered.

Class mailing list: multi_cse547a_wi22 at UW will be used sparingly by me mainly for announcements, but is open for posting by everyone on the list.

What will the course be about?
The course will be about methodological aspects of doing machine learning on big data. Big data allows our models to become gradually more complex, and this is an area called non-parametric statistics. A fundamental aspect of many non-parametric models is that they depend on the neighbors of a data point. Hence, search for neighbors, or similar items, in large data sets will be an important skill that we will develop.

Who is this class for?
This class is a core class in the Machine /Big Data PhD Track in Statistics. For any Statistics PhD student who wants to learn Machine Learning/Big Data, this class is part of the triplet of graduate courses 535 --> 538/548. Prerequisites are EITHER CSE 546 (Machine Learning) or STAT 535 (Foundations of Machine Learning). Either prerequisites are accepted to enroll in any of the STAT 548A, B or CSE 547.
Capacity permitting, the class is open to other graduate students with an interest in statistics, algorithms and computing, in particular to students involved in Machine Learning and Big Data research across campus.

Prerequisites

Knowledge of basic computer science principles and skills, at a level sufficient to write a reasonably non-trivial computer program (e.g., CS332, CS373 or equivalent are recommended).

Good knowledge of Python or Java will be extremely helpful since several assignments will require the use of parallel computing.

Basic probability theory (any introductory probability course).

Familiarity with writing rigorous proofs.

Familiarity with algorithmic analysis (e.g., CS332/CS373; CS417/CS421 would be more than necessary).

Fundamentals of statistics: Maximum Likelihood Estimation, MAP estimation, priors, likelihood, Bayes' formula.

Calculus and linear algebra: partial derivatives, gradient, the chain rule, vectors and matrices, matrix multiplications, eigenvalues and eigenvectors, positive definite matrices.

Learning Goals

Working knowledge of the main algorithmic and statistical issues raised by big data in machine learning, and of the paradigms to work with them
Working knowledge of paradigms and algorithms for data reduction by unsupervised learning: dimension reduction and clustering.
Working knowledge of methods for finding similar items in large data sets.
Concepts and working knowledge in network analysis and graph data
Basic knowledge of parallel computing with big data.
Ability to express formally in writing and to comprehend formally written ML results
Ability to apply the knowledge gained in a data analysis project.

Syllabus

Introduction: learning Big Data -- what is different?
Basics of parallel programming for big data
Nearest neighbors in high dimensions Locality Sensitive Hashig (LSH)
Dimension reduction and manifold learning
Clustering
- -- parametric clustering (K-means, EM)
- -- non-parametric clustering
  - - Bayesian non-parametric Dirichlet Process Mixture Models
  - - mode-seeking methods (mean-shift)
  - - level set methods
Graph and network data anaylysis
- - PageRank and Personalized Page Rank
- - spectral clustering
- - community detection
- - network models
Information retrieval
Data streams
- - Thomson sampling, reservoir sampling

Format: The course will consist of two weekly lectures, a series of homework assignments, a few quizzes, and a project. The TA will offer optional Tutorials where they will go over certain basic topics in more depth. In particular, some of the classes in the first part of the course will be partly flipped, and for these I encourage you to participate in the TA office hour.

Grading:The grade is based on homework + quizzes (approx 75%), project (approx 20%), class participation (approx 5%). These percentages are approximative, and may change by up to 5%.
The grading policies described here apply to students in good standing; students who engage in misconduct will be reported to the office of Community Standards & Student Conduct. Their grades may be handled differently.

Assignments: Typically assignments will be posted on the web on Thursdays (may be switched to Tuesday) and will be due the next Thursday at the beginnining of class. More generally, homework will be due one week after it is posted, and we will post a homework approximately every week. Graded homeworks and solutions will be available about a week after the homework due date.
Submit each homework as a single .pdf file through Canvas (with the exception of programming assignments, which will be discussed below)
The assignments will consist of (1) programming assignments (typically, to implement a version or a special case of an algorithm presented in the lecture) to be done in the programming language of your choice and (2) problems or other questions, including proofs. The programming assignments will be split into two separate parts:
- Output and results must be included in the single .pdf file that you submit.
- The code is submitted electronically, and separately from the homework pdf. You should submit the code in one or more files, a zip archive, or any usual and convenient format. We may use MOSS to check code plagiarism.
  With the exception of generic libraries (like plotting, matrix functions) you must write your own code. In particular, you are not to use matlab, R or python code for machine learning that is available on the web or with the textbook(s).
- You are strongly recommended to use python for this class. Support will be given to students unfamiliar or beginning with python. If you are one of these student, please communicate with the instructor and TAs immediately.
Late homeworks will be accepted in exceptional circumstances. Please let us know in advance if you think you will be late.
Teamwork: Each class participant submits her/his homework individually. Unless explicitly allowed to do so, you are required to write your own code. Discussing homework questions is acceptable as long as hints or solutions are not asked for or given; for example, a discussion to clarify what a question requires, on the discussion board or elsewhere, is acceptable. Please list on your homework the names of all the people you discussed with.
Quizzes will be announced. A quiz will take 10-15 minutes at the beginning of class, and the grades for them will be lumped together with the homework part of the grade. We are considering ways to give the quizzes on-line.
Participation What counts as participation? During lectures: Ask questions in class (chat OK if class on-line). Reply to the questions I pose during lectures.
On the discussion board: Answer questions on marked as counting for participation asked by instructor, or TAs, or other students. Note: discussion about the homework typically does not qualify for participation, but there are exceptions. For example, if you find an error in the homework and are the first to point it out to us, congratulations! we consider that participation.
How much participation is enough? Once a week on average, either in class or on discussion board, is sufficient for full grade.
Note that class attendance by itself is not graded as participation, and is not required.
Project Each project will have the same data set as starting point. You will perform a task on this data set.
You will submit a report describing what you did, submit your code, and test your trained predictor(s) on a test set with hidden labels. Evaluation of the project is based partly on the report and partly on the test set results. Time permitting, we will have short presentations/Q\&A during the last lecture time. More details about the project will be announced around week 3.

Religious accomodation Washington state law requires that UW develop a policy for accommodation of student absences or significant hardship due to reasons of faith or conscience, or fororganized religious activities. The UW’s policy, including more information about how to requestan accommodation, is available at Religious Accommodations Policy. Accommodations must berequested within the first two weeks of this course using the Religious Accommodations Request Form.

The UW food pantry A student should never have to make the choice between buying food or textbooks. The UW Food Pantry helps mitigate the social and academic effects of campus food insecurity. They aim to lessen the financial burden of purchasing food by providing students withaccess to food and hygiene products at no-cost. Students can expect to receive 4 to 5 days’ worth ofsupplemental food support when they visit the Pantry. For information including operating hours,location, and additional food support resources visit The UW Food Pantry. They can be found onthe North side of West Campus’ Poplar Hall at the corner of Brooklyn Ave NE and 41st. Last modified: Fri Sep 21 11:20:19 PDT 2018