BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
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
DTSTAMP:20260522T150115Z
LOCATION:D165
DTSTART;TZID=America/Chicago:20181112T110000
DTEND;TZID=America/Chicago:20181112T113000
UID:submissions.supercomputing.org_SC18_sess161_ws_pmbsf119@linklings.com
SUMMARY:Benchmarking Machine Learning Methods for Performance Modeling of 
 Scientific Applications
DESCRIPTION:Preeti Malakar, Prasanna Balaprakash, Venkatram Vishwanat, Vit
 ali Morozov, and Kalyan Kumaran (Argonne National Laboratory)\n\nPerforman
 ce modeling is an important and active area of research in high-performanc
 e computing (HPC). It helps in better job scheduling and also improves ove
 rall performance of coupled applications. Sufficiently rich analytical mod
 els are challenging to develop, however, because of interactions between d
 ifferent node components, network topologies, job interference, and applic
 ation complexity. When analytical performance models become restrictive be
 cause of application dynamics and/or multicomponent interactions, machine-
 learning-based performance models can be helpful. While machine learning (
 ML) methods do not require underlying system or application knowledge, the
 y are efficient in learning the unknown interactions of the application an
 d system parameters empirically using application runs. We present a bench
 mark study in which we evaluate eleven machine learning methods for modeli
 ng the performance of four representative scientific applications that are
  irregular and with skewed domain configurations on four leadership-class 
 HPC platforms. We assess the impact of feature engineering, size of traini
 ng set, modern hardware platforms, transfer learning, extrapolation on the
  prediction accuracy, and training and inference times. We find that baggi
 ng, boosting, and deep neural network ML methods are promising approaches 
 with median R^2 values greater than 0.95, and these methods do not require
  feature engineering. We demonstrate that cross-platform performance predi
 ction can be improved significantly using transfer learning with deep neur
 al networks.\n\nTag: Benchmarks, Parallel Programming Languages, Libraries
 , and Models, Performance, Simulation\n\nRegistration Category: Workshop R
 eg Pass\n\nSession Chair: Steven A. Wright (University of York, England)\n
 \n
END:VEVENT
END:VCALENDAR