Please join us for a UW Data Science Seminar on Thursday, January 23rd from 4:30 to 5:20 p.m. PT. The seminar will feature Ariane Ducellier, a postdoctoral scholar at the Institute for Health Metrics and Evaluation (IHME) at the University of Washington.

The seminar will be held in Hitchcock Hall 132 – Campus Map.

“Raking methods and applications to health metrics”

Abstract Raking is widely used in survey inference to adjust observations in contingency tables to given marginals. We propose a statistical approach, and a corresponding optimization algorithm, that is able to handle uncertain observations and margins to propagate the uncertainty and results in an efficient uncertainty quantification of the posterior (raked) estimates. Empirical results show that the approach obtains, at the cost of a single solve, nearly the same uncertainty estimates as computationally intensive Monte Carlo techniques that pass thousands of observed and of marginal samples through the entire raking process.  In many real situations, prior information in the form of ordinal constraints is available, and the adjusted observations table after raking must satisfy these constraints in order to be interpretable. We propose a new raking method that allows one to incorporate ordinal constraints, which has nearly the same running time of the conventional raking method. We illustrate the proposed approach on mortality rate data.

Ariane Ducellier, PhD, is a Postdoctoral Scholar at the Institute for Health Metrics and Evaluation (IHME) at the University of Washington, working in the Mathematical Sciences and Computational Algorithms team. Dr. Ducellier completed her doctoral work in Earth Sciences at University of Washington, where her research involved detecting and classifying earthquake events in large datasets of ground motion recordings. She also holds a master in Statistics from University of Washington.

The 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 and arts along with methodological areas in data science, such as computer science, applied math and statistics. Our presenters come from all domain fields and include occasional external speakers from regional partners, governmental agencies and industry.

The 2024-2025 seminars will be held in person, and are free and open to the public.