BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Metric and variable selection for functional nonparametric regress
 ion models - Torben Sell (University of Edinburgh)
DTSTART:20250131T140000Z
DTEND:20250131T150000Z
UID:TALK226138@talks.cam.ac.uk
CONTACT:Qingyuan Zhao
DESCRIPTION:In this talk\, we will consider nonparametric regression model
 s with multiple functional covariates. The focus of the work is to identif
 y relevant variables and useful metrics for the functional covariates\, an
 d to efficiently estimate the regression function. The proposed method is 
 based on an extension of the Nadaraya-Watson estimator\, where a kernel fu
 nction is applied to a linear combination of distance measures\, each comp
 uted on individual covariates\, in combination with an adaptive thresholdi
 ng step on the kernel weights. This data-driven least squares cross-valida
 tion method can asymptotically remove irrelevant noise variables and selec
 t relevant metrics for the functional covariates\, as will be shown both b
 y theory and numerical examples.
LOCATION:Centre for Mathematical Sciences MR12\, CMS
END:VEVENT
END:VCALENDAR