BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Statistical Methods for Ambulance Fleet Management - Woodard\, DB 
 (Cornell University)
DTSTART:20140423T132000Z
DTEND:20140423T135500Z
UID:TALK52124@talks.cam.ac.uk
CONTACT:Mustapha Amrani
DESCRIPTION:We introduce statistical methods to address two problems arisi
 ng in the management of ambulance fleets: (1) predicting the distribution 
 of ambulance driving time between arbitrary start and end locations in a r
 oad network\; and (2) space-time forecasting of ambulance demand. These pr
 edictions are critical for deciding how many ambulances should be deployed
  at a given time and where they should be stationed\, which ambulance shou
 ld be dispatched to an emergency\, and whether and how to schedule ambulan
 ces for non-urgent patient transfers. We demonstrate the accuracy and oper
 ational impact of our methods using ambulance data from Toronto Emergency 
 Medical Services.\n\nFor driving time prediction the relevant data are Glo
 bal Positioning System (GPS) recordings from historical lights-and-sirens 
 ambulance trips. Challenges include the typically large size of the road n
 etwork and dataset (70\,000 network links and 160\,000 historical trips fo
 r Toronto)\; the lack of trips in the historical data that follow precisel
 y the route of interest\; and the strong temporal\, weather\, and other ef
 fects on driving time. We introduce a model of the travel time at the trip
  level\, as well as a computationally efficient procedure for two-stage es
 timation in that model. For space-time forecasting of demand we develop in
 teger time-series factor models and spatio-temporal mixture models\, which
  capture the complex weekly and daily patterns in demand as well as change
 s in the spatial demand density over time.\n\nRelated Links: http://people
 .orie.cornell.edu/woodard/publications.html - Links to Woodard's publicati
 ons on ambulance fleet management\n
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR