BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Random forests - Gérard Biau\, Université Pierre et Marie Curie
DTSTART:20120120T160000Z
DTEND:20120120T170000Z
UID:TALK35202@talks.cam.ac.uk
CONTACT:Richard Samworth
DESCRIPTION:Random forests are a scheme proposed by Leo Breiman in the 200
 0's\nfor building a predictor ensemble with a set of decision trees that g
 row in\nrandomly selected subspaces of data. Despite growing interest and 
  practical\nuse\, there has been little exploration of the statistical pro
 perties of\nrandom forests\, and little is known about the mathematical fo
 rces driving\nthe algorithm.  In this talk\, we offer an in-depth analysis
  of a random\nforests model suggested by Breiman in 2004\, which is very c
 lose to the\noriginal algorithm. We show in particular that the procedure 
 is consistent\nand adapts to sparsity\, in the sense that its rate of conv
 ergence depends\nonly on the number of strong features and not on how many
  noise variables\nare present.
LOCATION:MR12\, CMS\, Wilberforce Road\, Cambridge\, CB3 0WB
END:VEVENT
END:VCALENDAR