BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//wp-events-plugin.com//7.2.3.1//EN
TZID:Europe/Paris
X-WR-TIMEZONE:Europe/Paris
BEGIN:VEVENT
UID:6045@i2m.univ-amu.fr
DTSTART;TZID=Europe/Paris:20220516T140000
DTEND;TZID=Europe/Paris:20220516T150000
DTSTAMP:20241120T200725Z
URL:https://www.i2m.univ-amu.fr/evenements/laplace-approximations-and-baye
 sian-p-splines-in-nonparametric-location-scale-models-for-interval-censore
 d-data/
SUMMARY:Philippe Lambert (Université de Liège & Université catholique de
  Louvain (Belgique)): Laplace approximations and Bayesian P-splines in non
 parametric  location-scale models for interval-censored data
DESCRIPTION:Philippe Lambert: A double additive model for the conditional m
 ean and standard deviation in location-scale models with a nonparametric e
 rror distribution is proposed. The response is assumed continuous and poss
 ibly subject to right or interval-censoring.\nNonparametric inference from
  censored data in location-scale models has been studied by many authors\,
  but it generally focuses on the estimation of conditional location and ca
 n only deal with the estimation of the smooth effects of a very limited nu
 mber of covariates.\nAdditive models based on P-splines are preferred here
  for their excellent properties and the possibility to handle a large numb
 er of additive terms (Eilers and Marx 2002). They are used to specify the 
 joint effect of covariates on location and dispersion within the location-
 scale model. A nonparametric error distribution with a smooth underlying h
 azard function and fixed moments is assumed for the standardized error ter
 m.\nIn the absence of interval-censoring\, a location-scale model with a s
 mall number of additive terms and a quartile-constrained error density (in
 stead of the hazard here) was considered in Lambert (2013) to analyse inte
 rval-censored data\, with inference relying on a numerically demanding MCM
 C algorithm.\nIt is shown how Laplace approximations to the conditional po
 sterior of spline parameters can be combined to bring fast and reliable es
 timation of the linear and additive terms\, and provide a smooth estimate 
 of the underlying error hazard function under moment constraints. These ap
 proximations are the cornerstones in the derivation of the marginal poster
 iors for the penalty parameters and smoothness selection. The resulting es
 timation procedures are motivated using Bayesian arguments and shown to ow
 n excellent frequentist properties. They are extremely fast and can handle
  a large number of additive terms within a few seconds even with pure R co
 de.\nThe methodology is illustrated with the analysis of right- and interv
 al-censored income data in a survey.\n** REFERENCES **\nEilers\, P.H.C. an
 d Marx\, B.D. (2002). Generalized linear additive smooth structures. Journ
 al of Computational and Graphical Statistics\, 11: 758-783.\nLambert\, P. 
 (2013). Nonparametric additive location-scale models for interval censored
  data. Statistics and Computing\, 23: 75-90.\nLambert\, P. (2021). Fast Ba
 yesian inference using Laplace approximations in nonparametric double addi
 tive location-scale models with right- and interval-censored data. Computa
 tional Statistics and Data Analysis\, 154\, 107088. https://doi.org/10.101
 6/j.csda.2020.107088\n&nbsp\;
ATTACH;FMTTYPE=image/jpeg:https://www.i2m.univ-amu.fr/wp-content/uploads/2
 022/05/Philippe_Lambert.png
CATEGORIES:Séminaire,Statistique
LOCATION:Saint-Charles - FRUMAM  (2ème étage)\, 3 Place Victor Hugo\, Mar
 seille\, 13003\, France
X-APPLE-STRUCTURED-LOCATION;VALUE=URI;X-ADDRESS=3 Place Victor Hugo\, Marse
 ille\, 13003\, France;X-APPLE-RADIUS=100;X-TITLE=Saint-Charles - FRUMAM  (
 2ème étage):geo:0,0
END:VEVENT
BEGIN:VTIMEZONE
TZID:Europe/Paris
X-LIC-LOCATION:Europe/Paris
BEGIN:DAYLIGHT
DTSTART:20220327T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
END:DAYLIGHT
END:VTIMEZONE
END:VCALENDAR