BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20260929T044233EDT-0013eNb1L9@132.216.98.100
DTSTAMP:20260929T084233Z
DESCRIPTION:Suchi Saria\, PhD Assistant Professor of Computer Science\, Sta
 tistics\, and Health Policy\, Johns Hopkins University Scalable Joint Mode
 ls for Reliable Event Prediction: Application to Monitoring Adverse Events
  using Electronic Health Record Data Tuesday\, 7 February 2017 3:30 pm – 4
 :30 pm - Purvis Hall\, 1020 Pine Ave. West\, Room 24\n\nALL ARE WELCOME \n
 \nAbstract: Many life-threatening adverse events such as sepsis and cardia
 c arrest are treatable if detected early. Towards this\, one can leverage 
 the vast number of longitudinal signals---e.g.\, repeated heart rate\, res
 piratory rate\, blood cell counts\, creatinine measurements---that are alr
 eady recorded by clinicians to track an individual's health. Motivated by 
 this problem\, we propose a reliable event prediction framework comprising
  two key innovations. First\, we extend existing state-of-the-art in joint
 -modeling to tackle settings with large-scale\, (potentially) correlated\,
  high-dimensional multivariate longitudinal data. For this\, we propose a 
 flexible Bayesian nonparametric joint model along with scalable stochastic
  variational inference techniques for estimation. Second\, we use a decisi
 on-theoretic approach to derive an optimal detector that trades-off the co
 st of delaying correct adverse-event detections against making incorrect a
 ssessments. On a challenging clinical dataset on patients admitted to an I
 ntensive Care Unit\, we see significant gains in early event-detection per
 formance over state-of-the-art techniques. \n\nSEE PDF FOR MORE DETAILS\n
DTSTART:20170207T203000Z
DTEND:20170207T213000Z
SUMMARY:SEMINAR: Scalable Joint Models for Reliable Event Prediction: Appli
 cation to Monitoring Adverse Events using Electronic Health Record Data
URL:https://www.mcgill.ca/epi-biostat-occh/channels/event/seminar-scalable-
 joint-models-reliable-event-prediction-application-monitoring-adverse-even
 ts-using-265517
END:VEVENT
END:VCALENDAR