BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//jEvents 2.0 for Joomla//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
UID:f101400a1cc76e8dac3811261f2c2e3361
CATEGORIES:Events
SUMMARY:USACM Large-Scale TTA Early-Career Colloquium (virtual)
DESCRIPTION:USACM Large-Scale TTA Early-Career Colloquium (virtual)Adaptive Sampling fo
r Constrained Optimization Under UncertaintyBrendan Keith\nBrown University
\nAbstract\n\nStochastic optimization problems with deterministic constrai
nts commonly appear in machine learning, finance, and engineering applicati
ons. This talk presents an improved adaptive solution strategy for this imp
ortant class of problems. The aim is to decrease the computational cost whi
le maintaining an optimal convergence rate. The guiding principle is to adj
ust the batch size (or sample size) on the fly so that the error in the gra
dient approximation remains proportional to the error in the underlying opt
imization problem. After providing motivation and context, I will present n
ew adaptive sampling algorithms that simultaneously maintain optimal sample
efficiency and iteration complexity for risk-neutral and risk-averse optim
ization under uncertainty with deterministic constraints. I will then demon
strate the efficacy of these algorithms in multiple applications, drawing m
ainly from use cases found in engineering design. This talk will provide an
introduction to adaptive sampling that aims to be accessible to a broad au
dience as well as showcase ongoing work in collaboration with Lawrence Live
rmore National Laboratory and UT Austin.\nBiography\n\nBrendan Keith is an
Assistant Professor in the Division of Applied Mathematics at Brown Univers
ity in Providence, Rhode Island. His research interests are mainly related
to the modeling and simulation of problems arising in natural sciences and
engineering, focusing on numerical methods for partial differential equatio
ns, scientific machine learning, and PDE-constrained optimization. In 2018,
Brendan received his Ph.D. in Computational Science, Engineering, and Math
ematics from the Oden Institute for Computational Engineering and Sciences
at the University of Texas at Austin. He has held postdoctoral positions at
TU Munich, ICERM, and Lawrence Livermore National Laboratory.\nSponsored b
y USACM Technical Thrust Area on Large Scale Structural Systems and Optimal
Design.\nContact for information about the seminar: admin@usacm.org. \nReg
ister
DTSTAMP:20221209T114520
DTSTART;TZID=America/Chicago:20221005T130000
DTEND;TZID=America/Chicago:20221005T140000
SEQUENCE:0
TRANSP:OPAQUE
END:VEVENT
END:VCALENDAR