BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//wp-events-plugin.com//7.4.0.1//EN
TZID:Europe/Paris
X-WR-TIMEZONE:Europe/Paris
BEGIN:VEVENT
UID:358@isdm.umontpellier.fr
DTSTART;TZID=Europe/Paris:20250407T140000
DTEND;TZID=Europe/Paris:20250407T140000
DTSTAMP:20260828T094010Z
URL:https://isdm.umontpellier.fr/events/tbd-3/
SUMMARY:TBD
DESCRIPTION:Online\n\nMachine Learning in Montpellier\, Theory &amp\; Pract
 ice - Julie Alberge (Inria)\n\nWhen dealing with right-censored data\, whe
 re some outcomes are missing due to a limited observation period\, surviva
 l analysis —known as time-to-event analysis — focuses on predicting th
 e time until an event of interest occurs. Multiple classes of outcomes lea
 d to a classification variant: predicting the most likely event\, a less e
 xplored area known as competing risks . Classic competing risks models cou
 ple architecture and loss\, limiting scalability. To address these issues\
 , we design a strictly proper censoring-adjusted separable scoring rule\, 
 allowing optimization on a subset of the data because the evaluation is co
 nducted independently for each observation. The loss estimates outcome pro
 babilities and enables stochastic optimization for competing risks\, which
  we use for efficient gradient boosting trees. SurvivalBoost not only outp
 erforms 12 state-of-the-art models across several metrics on 4 real-life d
 atasets\, both in competing risks and survival settings\, but also provide
 s great calibration\, the ability to predict across any time horizon\, and
  faster computation times compared to existing methods.\n\nMachine Learnin
 g in Montpellier\, Theory &amp\; Practice
ATTACH;FMTTYPE=image/jpeg:https://isdm.umontpellier.fr/wp-content/uploads/
 2026/06/ml-mtp-gC78d5.png
CATEGORIES:ML MTP
END:VEVENT
BEGIN:VTIMEZONE
TZID:Europe/Paris
X-LIC-LOCATION:Europe/Paris
BEGIN:DAYLIGHT
DTSTART:20250330T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
END:DAYLIGHT
END:VTIMEZONE
END:VCALENDAR