Tutorials: Mondays (5-6, Hugh Dempster Pavilion 110, beginning January 7)
Instructor Office Hours: Wednesdays (5, ICICS 193, beginning January 9)
TA Office Hours: Fridays (10-11, ICICS X237, beginning January 18)
Instructor: Mark Schmidt.
Teaching Assistants: Ainaz Hajimoradlou, Xiaomeng Ju
Synopsis: This is a graduate-level course on machine learning, a field that focuses on using automated data analysis for tasks like pattern recognition and prediction. The course will move quickly and assumes a strong background in math and computer science as well as previous experience with statistics and/or machine learning. The class is intended as a continuation of CPSC 340/532M and it is strongly recommended that you take CPSC 340/532M first before enrolling in CPSC 540. Topics will (roughly) include large-scale machine learning, density estimation, probabilistic graphical models, deep learning, and Bayesian statistics.
Registration and Prerequisites: Graduate and undergraduate students from any department are welcome to take the class, provided that they satisfy the prerequisites. However, you can only register automatically if you are enrolled as a graduate student in CPSC, EECE, or STAT. If you are a graduate student from a different department (or are an undergraduate student satisfying these requirements), you can register by following the instructions here and submitting the prerequisites form here. Graduate students in CPSC/EECE/STAT also need to submit the prerequisites form before the add/drop deadline to stay enrolled.
CPSC 340/532M vs. CPSC 540: CPSC 340 and CPSC 540 are roughly structured as one full-year course. CPSC 340 (which is also listed as CPSC 532M for graduate students) covers more data mining methods and the methods that are most widely-used in applications of machine learning while CPSC 540 covers more research-level machine learning methods and theory. It is strongly recommended that you take CPSC 340 first, as it covers the most common and practically-useful techniques. If CPSC 340 is full, you should still sign up for the CPSC 340 waiting list (not CPSC 540) as we may expand the class size: taking CPSC 540 because CPSC 340 is full is a terrible idea. In 540 it will be assumed that you are familiar with the material in the current offering of CPSC 340, and note that the Coursera machine learning course is not an adequate replacement for CPSC 340.
CPSC 540 requires a stronger computer science and math background and will require substantially more work (including proofs and implementing methods from scratch). Note that CPSC grad students typically only take 1-3 courses per term compared to 3-6 for undergraduate students: so you should expect the workload to be up to 3 times higher than in typical courses. If you want an introduction to machine learning, do not have a strong computer science and math background, or are mainly interested in applying machine learning in your research, then CPSC 340 is the right course to take. You can always decide to take (or audit) CPSC 540 later.
Auditting: Rather than registering as a student, an alternate option is to register as an auditor. This is a good option for students that may be missing some of the prerequisites or that don't have enough time to do all course requirements, but that still want exposure to the material. For graduate students, the form for auditing the course is available here. For undergraduates, you need to fill out the form here and indicate on the course information section that you wish to "audit". I will describe the auditting requirements and sign these forms on the first day of class.
Textbook: There is no textbook for the course, but the textbook with the most extensive coverage of many of the course's topics is Kevin Murphy's Machine Learning: A Probabilistic Perspective (MLAPP). This book can be purchased from Amazon, is on reserve in the CS Reading Room (ICCS 262), and can be accessed through the library here. Optional readings will be given out of this textbook, in addition to other free online resources.
Grading: Assignments 40%, Final 30%, Project 30%.
Piazza will be used for course-related questions.
|Date||Lecture Slides||Related Readings and Links||Homework|
|Wed Jan 2|| Syllabus
|MLAPP 1.1-1.2, 1.4, 6.5, 7.1-3, 7.5, 8.1-3
ML vs. Stats (2001, 2015) 3 Cultures of ML
Essence of Linear Algebra Mathematics for Machine Learning
|Fri Jan 4||Fundamentals of Learning||MLAPP 1.4, 6.5, Probability Primer|
|Mon Jan 7||Convex Optimization||MLAPP 7.4, 14.5, BV 2.1-2.3, 3.1-3.2 Taylor Polynomial|
|Wed Jan 9||Gradient Descent Convergence||BV 9.1-3|
|Fri Jan 11||Rates of Convergence||PL Inequality||Assignment 1 due|
|Mon Jan 14||Subgradients||MLAPP 13.3|
|Wed Jan 16||Proximal Gradient||MLAPP 13.5|
|Fri Jan 18||Structured Regularization||Structured Sparsity|| Assignment 2|
|Mon Jan 21||Coordinate Optimization||BV 9.4, Coordinate Descent|
|Wed Jan 23||Stochastic Subgradient||MLAPP 8.5, 13.4|
|Fri Jan 25||SGD Convergence Rate||Non-convex rates|
|Mon Jan 28||Stochastic Average Gradient||SAG|
|Wed Jan 30||Density Estimation||MLAPP 2.3, Covariance Matrix|
|Thu Jan 31||Guest Lecture: Richard Socher|
|Fri Feb 1||Multivariate Gaussians||MLAPP 2.4-5 and 4.1-3, Properties of Gaussians||Assignment 3 due|
|Mon Feb 4||Mixture Models||MLAPP 11.1-2|
|Wed Feb 6||Expectation Maximization||MLAPP 11.3-4, 11.6|
|Fri Feb 8||Kernel Density Estimation||MLAPP 14.7|| Assignment 3
|Mon Feb 11||Markov Chains||MLAPP 17.1-2|
|Wed Feb 13||Monte Carlo Methods||MLAPP 23.1-2|
|Fri Feb 15||Message Passing||MLAPP 17.4|
|Fri Feb 22||Reading Week||Assignment 3 due|
|Mon Mar 25||Hidden Markov Models||MLAPP 17.3-5, 18.1-4||Assignment 4
|Wed Mar 27||DAG Models||MLAPP 10.1-2, 10.5|
|Fri Mar 1||More DAGs||MLAPP 10.3-4, 26.1-4|
|Mon Mar 4||Undirected Graphical Models||MLAPP 19.1-4, 20.1-4|
|Wed Mar 6||Approximate Inference||MLAPP 24.1-2|
|Fri Mar 8||Log-Linear Models||MLAPP 19.5|
|Mon Mar 11||Boltzmann Machines||MLAPP 27.7, 28.1-2|
|Wed Mar 13||Conditional Random Fields||MLAPP 19.6|
|Fri Mar 15|| Guest Lecture: Mike Gelbart
Another look at CNNs
|Assignment 4 due|
|Mon Mar 18||Deep Structured Models|
|Wed Mar 20|| Fully-Convolutional Networks
Recurrent Neural Networks
|Fri Mar 22|| Long Short Term Memory
|MLAPP 3.1-4, 4.4-6, 5.1-4, 7.6||Assignment 5
|Mon Mar 25||Empirical Bayes||MLAPP 5.6-7|
|Wed Mar 27||Hierarchical Bayes||MLAPP 5.5|
|Fri Mar 29||Topic Models||MLAPP 27.1, 27.3|
|Mon Apr 1||More Approximate Inference||MLAPP 21.1-5, 22.1-2|
|Wed Apr 3|| Non-Parametric Bayes
VAEs and GANs
|MLAPP 15.1-3, 25.2|
Notes: These are additional notes mentioned in the lectures and homeworks:
Some related courses that have online notes are: