BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20190719T085743Z
LOCATION:HG F 1
DTSTART;TZID=Europe/Stockholm:20190612T160000
DTEND;TZID=Europe/Stockholm:20190612T163000
UID:submissions.pasc-conference.org_PASC19_sess145_msa256@linklings.com
SUMMARY:Parallelized High-Dimensional Sparse Inverse Covariance Matrix Est
 imation
DESCRIPTION:Minisymposium\nComputer Science and Applied Mathematics, Chemi
 stry and Materials, Physics, Engineering\n\nParallelized High-Dimensional 
 Sparse Inverse Covariance Matrix Estimation\n\nEftekhari, Bollhöfer, Schen
 k\n\nWe consider the problem of estimating sparse inverse covariance matri
 ces for high-dimensional datasets using the l1-regularized Gaussian maximu
 m likelihood method. This problem is particularly challenging as the requi
 red computational resources increase superlinearly with the number of rand
 om variables. We introduce a performant and scalable algorithm which build
 s on the current advancements of second-order methods. The routine leverag
 es the intrinsic parallelism in the linear algebra operations and exploits
  the underlying sparsity of the problem. The computational bottlenecks are
  identified and the respective subroutines are parallelized using a hybrid
  MPI-OpenMP approach. Numerical examples conducted at the Swiss National S
 upercomputing Center (Cray XC40) show that, in comparison to the state-of-
 the-art algorithms, the proposed routine provides significant speed-up wit
 h scalability up to 128 nodes. The developed framework is used to approxim
 ate the sparse inverse covariance matrix for datasets with up to 10 millio
 n random variables.
END:VEVENT
END:VCALENDAR