[CS] Reminder: [masters-presentation] Barbarioli/MS Presentation/Nov 25, 2020

[Tricia Baclawski]

This is an announcement of Bruno Barbarioli's MS Presentation.


https://uchicago.zoom.us/j/97178502911?pwd=WW5SbUh5OFlUTTEyamJqSENhNWM3UT09
Password: 741561

------------------------------------------------------------------------------
Date:  Wednesday, November 25, 2020

Time:  9:00 AM

Place:  via zoom

M.S. Candidate:  Bruno Barbarioli

M.S. Paper Title: Adaptive Online Block Subsampling for Time Series
Data

Abstract:
Time series data are ubiquitous in modern day applications ranging
from IoT to finance. The ever increasing volume of data, however,
could make its analysis infeasible in some cases. In this scenario,
classical database subsampling techniques are ineffective because they
can break up important temporal correlations. We propose a new general
framework for subsampling in blocks (or contiguous observations),
rather than individual time-points, with an adaptive stopping criteria
that determines when it has sampled enough data to accurately
represent a distribution. The framework is composed of a breakpoint
detector that decomposes time-series into stationary segments and
applies a consistent subsampling algorithm within segments. Our method
has the following desirable properties: (1) it captures temporal
correlations in a more data-efficient way than uniform sampling, (2)
it adaptively determines how much to sample and when, (3) it can be
directly applied to delta-encoded data if the block size is chosen
appropriately. Our experiments suggest that our method is superior to
every other currently in use when it comes to capturing the dependence
structure of a time series, in particular the autocorrelation between
data points and estimated coefficients of ARIMA processes.

Bruno's advisor is Prof. Sanjay Krishnan

Login to the Computer Science Department website for details:
 https://newtraell.cs.uchicago.edu/phd/ms_announcements#barbarioli

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Tricia Baclawski
Student Affairs Administrator
Computer Science Department
5730 S. Ellis - Room 350
Chicago, IL 60637
pbaclawski at uchicago.edu
(773) 702-6854
/pronouns: she, her, hers/
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=