BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//jEvents 2.0 for Joomla//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VTIMEZONE
TZID:Europe/Budapest
END:VTIMEZONE
BEGIN:VEVENT
UID:3mmckjrc81snnoda72aqih97rg@google.com
CATEGORIES:{lang hu}Sztochasztika szeminárium{/lang}{lang en}Stochastics seminar{/lang}
SUMMARY:Szabó Botond (Universiteit Leiden): On the fundamental understanding of distributed computation
LOCATION:Szeged, Aradi vértanúk tere 1., Riesz terem
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:Abstract: In recent years, the amount of available information has become s
o vast in certain fields of applications that it is infeasible or undesirab
le to carry out all the computations on a single server. This has motivated
the design and study of distributed statistical or learning approaches. In
distributed methods, the data is split amongst different administrative un
its and computations are carried out locally in parallel to each other. The
outcome of the local computations are then aggregated into a final result
on a central machine.

In this talk we will consider the limitations and
guarantees of distributed methods under communication constraints (i.e. on
ly limited amount of bits are allowed to be transmitted between the machine
s) in context of the random design regression model. We derive minimax lowe
r bounds, matching upper bounds and provide adaptive estimators reaching th
ese limits.

This is a joint work with Harry van Zanten.
DTSTAMP:20200809T075010Z
DTSTART;TZID=Europe/Budapest:20190920T140000
DTEND;TZID=Europe/Budapest:20190920T160000
SEQUENCE:1
TRANSP:OPAQUE
END:VEVENT
END:VCALENDAR