Relatore:
Ghislaine GAYRAUD
- Université de Technologie, Compiègne
mercoledì 1 marzo 2017
alle ore
12.30
Polo Santa Marta, Via Cantarane 24, Room 1.59
We derive posterior concentration rates in a bivariate nonparametric quantile regression model with the resort of the Asymmetric Laplace Distribution; the latter leads to a misspecified inferential problem. Applying our general results to the Sobolev and Hôlder class along with a specific prior, we retrieve the well-known optimal (minimax) rates of convergence only for the Sobolev class. We illustrate our theoretical results by simulations.