NonpModelCheck: An R Package for Nonparametric Lack-of-Fit Testing and Variable Selection

Autor(es)

Zambom, Adriano Zanin ; Akritas, Michael G.

• Tipo de documento: Artigo Científico
• Fonte: JOURNAL OF STATISTICAL SOFTWARE; v. 77, n. 10, p. 1-28, APR 2017.
• Citações Web of Science: 1

Resumo

We describe the R package NonpModelCheck for hypothesis testing and variable selection in nonparametric regression. This package implements functions to perform hypothesis testing for the significance of a predictor or a group of predictors in a fully nonparametric heteroscedastic regression model using high-dimensional one-way ANOVA. Based on the p values from the test of each covariate, three different algorithms allow the user to perform variable selection using false discovery rate corrections. A function for classical local polynomial regression is implemented for the multivariate context, where the degree of the polynomial can be as large as needed and bandwidth selection strategies are built in. (AU)

• Processo FAPESP: 12/10808-2 - Algoritmos para Testes de Hipótese em Regressão Não-Paramétrica e suas Propriedades Assintóticas com Aplicações em Seleção de Variáveis
• Beneficiário: Adriano Zanin Zambom
• Modalidade de apoio: Bolsas no Brasil - Pós-Doutorado