Directory Calendar News Careers Alumni Giving

Biostatistics Seminars

Department of Biostatistics Seminar Series

The Department of Biostatistics presents a regular speaker series each semester, generally on Thursday afternoon each week. Diverse experts lecture on their work in biostatistics.

Upcoming Biostats Seminars

Thu 3/1/2018 3:30PM - 4:30PM
Inference and Variable Selection with Random Forests - Lucas Mentch
Public Health Auditorium (G23)

Biostatistics guest speaker, Lucas Mentch, University of Pittsburgh, Statistics, will present, "Inference and Variable Selection with Random Forests."
Thu 4/5/2018 3:30PM - 4:30PM
Seonjoo Lee, Columbia University
Public Health Auditorium (G23)

Biostatistics guest speaker, Seonjoo Lee, Columbia University, will present.
Thu 4/12/2018 3:30PM - 4:30PM
Ciprian Crainiceanu, Johns Hopkins University
Public Health Auditorium (G23)

Biostatistics guest speaker, Ciprian Crainiceanu, Johns Hopkins University, will present.
Thu 4/19/2018 3:30PM - 4:30PM
Wensheng Guo, University of Pennsylvania
Public Health Auditorium (G23)

Biostatistics guest speaker, Wensheng Guo, University of Pennsylvania, will present.

Previous Biostats Seminars

Biostatistics Seminar Series

Biostatistics Seminar: David Choi - Carnegie Mellon University

Thursday 9/28 3:30PM - 5:00PM
Public Health Auditorium (G23)

Biostatistics Seminar guest speaker, David Choi, Carnegie Mellon University, will present, "Estimation of Monotone Treatment Effects in Network Experiments".

Randomized experiments on social networks pose statistical challenges, due to the possibility of interference (such as peer influence) between units. To find rigorous confidence intervals on the average treatment effect in such settings, one typically must model the underlying social network --  "who can influence whom", how such effects might combine, and whether they can cascade over long distances. In many settings, this may be an unreasonable modeling burden. As an alternative, we propose new methods for finding confidence intervals on the attributable treatment effect. These methods do not make assumptions on the structure of the underlying social network, but instead require an identifying assumption that is similar to requiring nonnegative treatment effects; for example, assuming that a vaccine does not increase the risk of catching a disease, either directly or indirectly through vaccinated peers. Network or spatial information can be used to customize the test statistic; in principle, this can increase power to detect spillovers without making further assumptions on the data generating process.

Last Updated On Tuesday, August 29, 2017 by Haydo, Amber LC
Created On Wednesday, August 09, 2017


For information on seminars and events in the department, contact:

Bernadette Kapko

© 2018 by University of Pittsburgh Graduate School of Public Health

Login  |  Sitemap