BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20260415T215410EDT-5141mmrd28@132.216.98.100
DTSTAMP:20260416T015410Z
DESCRIPTION:Title: Nonparametric location-scale models for right- and inter
 val-censored data with inference based on Laplace approximations.\n\nAbstr
 act: A double additive model for the conditional mean and standard\n	\n	devi
 ation in location-scale models with a nonparametric error distribution is 
 proposed. The response is assumed continuous and possibly subject to right
  or interval-censoring. Nonparametric inference from censored data in loca
 tion-scale models has been studied by many authors\, but it generally focu
 ses on the estimation of conditional location and can only deal with the e
 stimation of the smooth effects of a very limited number of covariates. Ad
 ditive models based on P-splines are preferred here for their excellent pr
 operties and the possibility to handle a large number of additive terms (E
 ilers and Marx 2002). They are used to specify the joint effect of covaria
 tes on location and dispersion within the location-scale model. A nonparam
 etric error distribution with a smooth underlying hazard function and fixe
 d moments is assumed for the standardized error term. In the absence of ri
 ght-censoring\, a location-scale model with a small number of additive ter
 ms and a quartile-constrained error density (instead of the hazard here) w
 as considered in Lambert (2013) to analyse interval-censored data\, with i
 nference relying on a numerically demanding MCMC algorithm. It is shown ho
 w Laplace approximations to the conditional posterior of spline parameters
  can be combined to bring fast and reliable estimation of the linear and a
 dditive terms\, and provide a smooth estimate of the underlying error haza
 rd function under moment constraints. These approximations are the corners
 tones in the derivation of the marginal posteriors for the penalty paramet
 ers and smoothness selection. The resulting estimation procedures are moti
 vated using Bayesian arguments and shown to own excellent frequentist prop
 erties. They are extremely fast and can handle a large number of additive 
 terms within a few seconds even with pure R code. The methodology is illus
 trated with the analysis of right- and interval-censored income data in a 
 survey.\n\n \n\n \n
DTSTART:20211130T153000Z
DTEND:20211130T163000Z
SUMMARY:Philippe LAMBERT (Université de Liège et Université catholique de L
 ouvain\, Belgique)
URL:https://www.mcgill.ca/mathstat/channels/event/philippe-lambert-universi
 te-de-liege-et-universite-catholique-de-louvain-belgique-335169
END:VEVENT
END:VCALENDAR