A method for variable selection and structure discovery in the contextof nonparametric regression in high dimensions is proposed in a forthcoming paper, where a small subset of variables are relevant and may have nonlinear effects on theresponse. The proposed method, called the GRID, is an extension of the RODEO method of Lafferty & Wasserman (2008), which only makes variable selection. In this paper we briefly describe the method and present the main theoretical founda-tions of the two stages of the procedure: (i) variable selection with linear/nonlinearclassification of the covariates and (ii) identification of interactions.
STRUCTURE DISCOVERING IN NONPARAMETRIC REGRESSION BY THE GRID PROCEDURE
Francesco Giordano;Soumendra Nath Lahiri;Maria Lucia Parrella
2019-01-01
Abstract
A method for variable selection and structure discovery in the contextof nonparametric regression in high dimensions is proposed in a forthcoming paper, where a small subset of variables are relevant and may have nonlinear effects on theresponse. The proposed method, called the GRID, is an extension of the RODEO method of Lafferty & Wasserman (2008), which only makes variable selection. In this paper we briefly describe the method and present the main theoretical founda-tions of the two stages of the procedure: (i) variable selection with linear/nonlinearclassification of the covariates and (ii) identification of interactions.File in questo prodotto:
Non ci sono file associati a questo prodotto.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
