Course information
Instructors: Rina Foygel Barber (UChicago) and Will Fithian (UC Berkeley).
Course description: Selective inference means searching for interesting patterns in data, with statistical guarantees that account for the search process. It encompasses multiple testing, post-selection inference, and adaptive or interactive inference. The course will emphasize recent advances and open problems in the field including knockoffs, multiple testing in structured problems, and conditional post-selection inference.
This PhD topics course is cross-listed at the University of Chicago and UC Berkeley. The course will begin on Wednesday September 30, 2020, the second day of the UChicago Fall term.
Schedule: MW, 9am-10am Pacific time / 11am-12pm Central time, plus the occasional Friday:
On the first week, there is no class on Monday so we will instead meet on Wed+Fri
Wed Nov 11 is a day off at Berkeley, so that week we will instead hold class on Mon+Fri
Logistics: Each week, students will watch one video lecture by one of the course instructors, and one video seminar from the International Seminar on Selective Inference (ISSI). Scheduled course time will be used for discussion - typically, the lecture will be discussed on Mondays, and the ISSI seminar will be discussed on Wednesdays. We will also use Piazza for discussions and Q&A.
Participation: Since this is a discussion-based course, students are expected to attend all classes (an occasional absence is of course fine if needed). Students are expected to participate actively both in the live discussions and in the discussions on Piazza.
Project: Each student is expected to complete a research project that is related in some way to the themes and topics of the course. Projects may be done individually or in small groups. Students are welcome to choose a project that is integrated with their thesis research. The instructors will discuss project ideas and progress with students individually.