BEGIN:VCALENDAR VERSION:2.0 PRODID:-//wp-events-plugin.com//7.3.1//EN TZID:America/Los_Angeles X-WR-TIMEZONE:America/Los_Angeles BEGIN:VEVENT UID:72@escience.washington.edu DTSTART;TZID=America/Los_Angeles:20240215T163000 DTEND;TZID=America/Los_Angeles:20240215T172000 DTSTAMP:20240212T173358Z URL:https://escience.washington.edu/events/uwdss-mccormick/ SUMMARY:UW Data Science Seminar: Tyler McCormick DESCRIPTION:\n\nPlease join us for a UW Data Science Seminar on Thursday\, February 15th from 4:30 to 5:20 p.m. PST. The seminar will feature Tyler M cCormick\, a Senior Data Science Fellow and a Professor of Statistics and Sociology at UW.\n\nThe seminar will be held in the Electrical &\; Comp uter Engineering Building (ECE)\, Room 105\n\n\n \;\n"Robustly estimat ing heterogeneity in factorial data using Rashomon Partitions"\nAbstract: Many statistical analyses begin with a fundamental question: How does the outcome vary with observable covariates?  Do more experienced employees r eceive higher wages? Do vaccinated individuals get sick less frequently th an those who are unvaccinated? Do trout in warm water eat less than trout in cold water? Does Venus's atmosphere contain a higher fraction of nitrog en than Mars?  Do patients taking Metformin have lower A1C measures than those taking DPP4-inhibitors?\n\nIn settings where these covariates are di screte\, this problem yields a factorial-like structure\, and it is imposs ible to enumerate all possible combinations of covariates for any scientif ically interesting setting. In this paper\, we propose an approach to enum erating heterogeneity in the relationship between an outcome and discrete covariates by creating a Rashomon Partitions Set (RPS). Each Rashomon part ition consists of the feature combinations that maximize heterogeneity in the outcome space. We construct this by pooling similar feature combinati ons using priors over pooling patterns in an overarching Bayesian model.  We show that we can characterize the set of Rashomon Partitions in terms of its fraction of the overall posterior and size.  Further\, we demonstr ate that the RPS is enumerable in meaningful settings by leveraging the in sight that many potential combinations of features are\, in practice\, non sensical for pooling because they represent different dimensions in the co variate space.  We demonstrate RPS construction in the context of two pra ctical settings: finding heterogeneity in outcomes of a randomized trial a nd examining racial disparities in health outcomes in a large clinical dat aset.  This is joint work with Arun Chandrasekhar (Stanford Economics) an d Aparajithan Venkateswaran (UW Statistics).\n\n\nBio: Tyler McCormick is a Professor of Statistics and Sociology at the University of Washington\, where he is also a core faculty member in the Center for Statistics and th e Social Sciences.  He is also a Senior Data Science Fellow at the eScien ce Institute.  Tyler's work develops statistical models that infer depend ence structure in scientific settings where data are sparsely observed or observed subject to error.  His recent projects include estimating featur es of social networks (e.g. the degree of clustering or how central an ind ividual is) using data from standard surveys\, inferring a likely cause of death (when deaths happen outside of hospitals) using reports from surviv ing caretakers\, and quantifying &\; communicating uncertainty in predi ctive models for global health policymakers.  He holds a Ph.D. in Statist ics (with distinction) from Columbia University and is the recipient of an NIH New Innovator (DP2) Award\, NIH Career Development (K01) Award\, Army Research Office Young Investigator Program Award\, and a Google Faculty R esearch Award.  Tyler is the former Editor of the Journal of Computation al and Graphical Statistics (JCGS) and a Fellow of the American Statistic al Association.\nThe UW Data Science Seminar is an annual lecture series at the University of Washington that hosts scholars working across applied areas of data science\, such as the sciences\, engineering\, humanities a nd arts along with methodological areas in data science\, such as computer science\, applied math and statistics. Our presenters come from all domai n fields and include occasional external speakers from regional partners\, governmental agencies and industry.\nThe 2023-2024 seminars will be held in person\, and are free and open to the public. LOCATION:Electrical and Computer Engineering Building\, Room 105\, 185 W St evens Way NE\, Seattle\, WA\, 98195\, United States X-APPLE-STRUCTURED-LOCATION;VALUE=URI;X-ADDRESS=185 W Stevens Way NE\, Seat tle\, WA\, 98195\, United States;X-APPLE-RADIUS=100;X-TITLE=Electrical and Computer Engineering Building\, Room 105:geo:0,0 END:VEVENT BEGIN:VTIMEZONE TZID:America/Los_Angeles X-LIC-LOCATION:America/Los_Angeles BEGIN:STANDARD DTSTART:20231105T010000 TZOFFSETFROM:-0700 TZOFFSETTO:-0800 TZNAME:PST END:STANDARD END:VTIMEZONE END:VCALENDAR