BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/New_York
X-LIC-LOCATION:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20210402T160551Z
LOCATION:Track 8
DTSTART;TZID=America/New_York:20201111T154000
DTEND;TZID=America/New_York:20201111T154500
UID:submissions.supercomputing.org_SC20_sess198_ws_whpc103@linklings.com
SUMMARY:A Machine Learning Classifier of Damaging Earthquakes as a Microse
 rvice in the Urgent Computing Workflow
DESCRIPTION:Workshop\n\nA Machine Learning Classifier of Damaging Earthqua
 kes as a Microservice in the Urgent Computing Workflow\n\nMonterrubio Vela
 sco, Carrasco, Rojas, Rodríguez, Modesto...\n\nThe package Urgent Computin
 g Integrated Services for EarthQuakes (UCIS4EQ) is a workflow developed in
  the framework of the Center of Excellence for Exascale in Solid Earth (Ch
 EESE). Inside this complex system, an important task of UCIS4EQ involves p
 redicting the damage potential of a new earthquake in order to provide inf
 ormation for the decision-maker to habilitate the protocol, guidelines, an
 d the computing resources to run an urgent job in HPC centers. In this wor
 k, we present a method that classifies the damage potential of an earthqua
 ke by using machine learning algorithms. Our method requires the informati
 on of the estimates of peak ground accelerations (PGA, at 90% non-exceedan
 ce for 50 years); socio-economic characteristics of the affected countries
 ; and the historical Significant Earthquake Database.  As a target, we use
  the modified Mercalli intensity (MMI) scale. We consider the PGA estimate
 s, latitude, longitude, depth, magnitude, infrastructure index and infrast
 ructure quality as features. As classifier algorithms, we apply the Random
  Forest, Support Vector Machine, and XGBoost. Our result is a metamodel th
 at uses a voting system to classify an event as urgent or non-urgent. Our 
 results indicate an accuracy of approximately 70 percent. We consider such
  accuracy satisfactory given the short training time, the novelty of the m
 ethodology and the database quality.  The code developed in this work will
  provide a microservice in UCIS4EQ, and thus, it will assist in the urgent
  seismic simulation developed in ChEESE.\n\nTag: Education, Training and O
 utreach, Professional Development, Workforce Development\n\nRegistration C
 ategory: Workshop Reg Pass
END:VEVENT
END:VCALENDAR