BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Denver
X-LIC-LOCATION:America/Denver
BEGIN:DAYLIGHT
TZOFFSETFROM:-0700
TZOFFSETTO:-0600
TZNAME:MDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0600
TZOFFSETTO:-0700
TZNAME:MST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20200129T163603Z
LOCATION:501
DTSTART;TZID=America/Denver:20191117T163000
DTEND;TZID=America/Denver:20191117T165500
UID:submissions.supercomputing.org_SC19_sess103_ws_indis104@linklings.com
SUMMARY:Training Classifiers to Identify TCP Signatures in Scientific Work
 flows
DESCRIPTION:Workshop\n\nTraining Classifiers to Identify TCP Signatures in
  Scientific Workflows\n\nPapadimitriou, Kiran, Wang, Mandal, Deelman\n\nId
 entifying network anomalies is an important measure to ensure reliability 
 and quality of data transfers among facilities. Scientific workflows in pa
 rticular heavily rely on good network performance to ensure their smooth e
 xecutions. In this paper, we present a lightweight classifier system that 
 is able to recognize anomalous TCP transfers. Using random forest trees an
 d labeled data sets, we evaluate the classifier with real workflow transfe
 rs for ground truth data. Our studies reveal that various TCP congestion a
 lgorithms behave differently in anomalous conditions. We show that trainin
 g classifiers on these separately can aid detection in network performance
  deterioration. Results reveal that our classifiers are able to better pre
 dict anomalous flows for TCP Reno and Hamilton compared to Cubic and BBR, 
 due to the manner how their congestion control algorithms handle the anoma
 lies.\n\nTag: Workshop Reg Pass, Big Data, Data Analytics, Datacenter, Net
 works, Software-defined networking\n\nRegistration Category: Workshop Reg 
 Pass, Big Data, Data Analytics, Datacenter, Networks, Software-defined net
 working
URL:https://sc19.supercomputing.org/presentation/?id=ws_indis104&sess=sess
 103
END:VEVENT
END:VCALENDAR