BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20260416T074537EDT-0136bf7diA@132.216.98.100
DTSTAMP:20260416T114537Z
DESCRIPTION:Title: Imbalanced learning using actuarial modified loss functi
 on in tree-based models\n\nAbstract:\n\nTree-based models have gained mome
 ntum in insurance claim loss modeling\; however\, the point mass at zero a
 nd the heavy tail of insurance loss distribution pose the challenge to app
 ly conventional methods directly to claim loss modeling. With a simple ill
 ustrative dataset\, we first demonstrate how the traditional tree-based al
 gorithm’s splitting function fails to cope with a large proportion of data
  with zero responses. To address the imbalance issue presented in such los
 s modeling\, this paper aims to modify the traditional splitting function 
 of Classification and Regression Tree (CART). In particular\, we propose t
 wo novel actuarial modified loss functions\, namely\, the weighted sum of 
 squared error and the sum of squared Canberra error. These modified loss f
 unctions impose a significant penalty on grouping observations of non-zero
  response with those of zero response at the splitting procedure\, and thu
 s significantly enhance their separation. Finally\, we examine and compare
  the predictive performance of such actuarial modified tree-based models t
 o the traditional model on synthetic datasets that imitate insurance loss.
  The results show that such modification leads to substantially different 
 tree structures and improved prediction performance.\n\n\n	Speaker\n\n\nZhi
 yu (Frank) Quan is an Assistant Professor at the Department of Mathematics
  of the University of Illinois at Urbana-Champaign. He holds a Ph.D. in Ac
 tuarial Science from the University of Connecticut. Before joining Illinoi
 s\, he worked for a cutting-edge Insurtech company as a R & D data scienti
 st developing data-driven solutions for major insurance companies. He has 
 a broad spectrum of research interests in data science applications in ins
 urance such as tree-based models\, natural language processing\, deep lear
 ning\, and applies his actuarial expertise to build predictive models for 
 claim research\, rate making\, etc.\n\n \n\nhttps://mcgill.zoom.us/j/83436
 686293?pwd=b0RmWmlXRXE3OWR6NlNIcWF5d0dJQT09\n\nMeeting ID: 834 3668 6293\n
 \nPasscode: 12345\n\n\n	\n		\n			 \n		\n	\n\n
DTSTART:20211008T193000Z
DTEND:20211008T203000Z
SUMMARY:Zhiyu Quan (University of Illinois at Urbana-Champaign)
URL:https://www.mcgill.ca/mathstat/channels/event/zhiyu-quan-university-ill
 inois-urbana-champaign-333987
END:VEVENT
END:VCALENDAR