Adaptive Estimation in an Autoregression and a Geometrical B-Mixing Regression Framework

We study the problem of estimating some unknown regression or autoregression function in a B-mixing dependent framework (for the design or the errors). For this end, we consider some collection of models which are finite dimensional spaces. A penalized least-squares estimator (PLSE) is built on a data driven selected model among this collection. We provide non asymptotic risk bounds for this PLSE from which we deduce adaptivity properties.

MoreLess

Year of publication:	1998
Authors:	Baraud, Y. ; Comte, F. ; Viennet, G.
Subject:	MODELS \| LEAST SQUARE \| EVALUATION

More details

Type of publication:	Book / Working Paper
Notes:	34 pages
Classification:	C13 - Estimation ; C14 - Semiparametric and Nonparametric Methods
Source:	RePEc - Research Papers in Economics

Persistent link: https://ebvufind01.dmz1.zbw.eu/10005671512