[Sds-seminars] Fwd: [YINS] Tomorrow at 12: Daniel Roy, "In Defense of Uniform Convergence: Generalization via derandomization with an application to interpolating predictors"

[Dan Spielman]

---------- Forwarded message ---------
From: Hau, Emily <emily.hau at yale.edu>
Date: Tue, Nov 17, 2020 at 2:59 PM
Subject: [YINS] Tomorrow at 12: Daniel Roy, "In Defense of Uniform
Convergence: Generalization via derandomization with an application to
interpolating predictors"
To: yins at mailman.yale.edu <yins at mailman.yale.edu>


*[image: cid:8CAC2F06-ABAC-4072-82E0-5A0A6FFE1226 at its.yale.internal]*
<https://yins.yale.edu/event/yins-seminar-daniel-roy-university-toronto>

YINS Seminar: Wednesday, November 18, 2020, 12:00-1:00pm
<https://yins.yale.edu/event/yins-seminar-daniel-roy-university-toronto>

*“In Defense of Uniform Convergence: Generalization via derandomization
with an application to interpolating predictors”*



*Speaker: Daniel Roy*
*Associate Professor, University of Toronto*

*J**oined by graduate student Jeffrey Negrea *

*To participate:*
Join from PC, Mac, Linux, iOS or Android: https://yale.zoom.us/j/95827054840
    Or Telephone：203-432-9666 (2-ZOOM if on-campus) or 646 568 7788
    Meeting ID: 958 2705 4840
    International numbers available: https://yale.zoom.us/u/aciY0peggr

*Abstract: *We propose to study the generalization error of a learned
predictor ^h in terms of that of a surrogate (potentially randomized)
predictor that is coupled to ^h and designed to trade empirical risk for
control of generalization error. In the case where ^h interpolates the
data, it is interesting to consider theoretical surrogate classifiers that
are partially derandomized or rerandomized, e.g., fit to the training data
but with modified label noise. We also show that replacing ^h by its
conditional distribution with respect to an arbitrary σ-field is a
convenient way to derandomize. We study two examples, inspired by the work
of Nagarajan and Kolter (2019) and Bartlett et al. (2019), where the
learned classifier ^h interpolates the training data with high probability,
has small risk, and, yet, does not belong to a nonrandom class with a tight
uniform bound on two-sided generalization error. At the same time, we bound
the risk of ^h in terms of surrogates constructed by conditioning and
denoising, respectively, and shown to belong to nonrandom classes with
uniformly small generalization error.

Joint work by Jeffrey Negrea, Gintare Karolina Dziugaite, Daniel M. Roy

For speaker bio, please visit http://danroy.org

*Upcoming Seminars:*


*December 2, 2020, 12:00pm YINS Distinguished Lecturer Seminar: Peter
Bartlett (UC Berkeley)
<https://yins.yale.edu/event/yins-distinguished-lecturer-seminar-peter-bartlett-uc-berkeley>*



Emily E. H. Hau | Director, Programs and Partnerships

*Yale Institute for Network Science*

*Yale University*

17 Hillhouse Avenue
<https://www.google.com/maps/search/17+Hillhouse+Avenue+%7C+Room+341+%7C+New+Haven,+CT+06511?entry=gmail&source=g>
 | Room 341
<https://www.google.com/maps/search/17+Hillhouse+Avenue+%7C+Room+341+%7C+New+Haven,+CT+06511?entry=gmail&source=g>
 |
<https://www.google.com/maps/search/17+Hillhouse+Avenue+%7C+Room+341+%7C+New+Haven,+CT+06511?entry=gmail&source=g>
 New Haven, CT 06511
<https://www.google.com/maps/search/17+Hillhouse+Avenue+%7C+Room+341+%7C+New+Haven,+CT+06511?entry=gmail&source=g>

c: (203) 273-7886

emily.hau at yale.edu






_______________________________________________
YINS mailing list
YINS at mailman.yale.edu
https://mailman.yale.edu/mailman/listinfo/yins
-- 
Sent from mobile phone
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.yale.edu/pipermail/sds-seminars/attachments/20201117/a586bb29/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.png
Type: image/png
Size: 83529 bytes
Desc: not available
URL: <http://mailman.yale.edu/pipermail/sds-seminars/attachments/20201117/a586bb29/attachment.png>