Portada > Escuela de Doctorado de Navarra - EDONA > Tablón de anuncios
Resumen:
Penalised splines are a general and versatile approach to estimate
nonlinear effects of continuous covariates based on penalised least
squares or penalized likelihood principles. In this presentation, we will introduce a mixed model representation of penalised spline smoothing as well as the corresponding Bayesian perspective. In addition, we will discuss how penalized splines can be used as building blocks in a modular approach to semiparametric regression including spatial- and spatio-temporal smoothing or the inclusion of individual-specific curves for longitudinal data.