Adriano Zanin Zambom's home page

NonpModelCheck: Model Checking and Variable Selection in Nonparametric Regression

Provides tests of significance for covariates (or groups of covariates) in a fully nonparametric regression model and a variable (or group) selection procedure based on False Discovery Rate. In addition, it provides a function for local polynomial regression for any number of dimensions, using a bandwidth specified by the user or automatically chosen by cross validation or an adaptive procedure.

Reference Manual: NonpModelCheck.pdf

Package source: NonpModelCheck_3.0.tar.gz

Paper in Journal of Statistical Software: NonpModelCheck: An R Package for Nonparametric Lack-of-Fit Testing and Variable Selection

SignifReg: Consistent Significance Controlled Variable Selection in Linear Regression

Provides significance controlled variable selection algorithms with different directions (forward, backward, stepwise) based on diverse criteria (AIC, BIC, adjusted r-square, PRESS, or p-value). The algorithm selects a final model with only significant variables based on a correction choice of False Discovery Rate, Bonferroni, or no correction.

Reference Manual: SignifReg.pdf

Package source: SignifReg_3.0.tar.gz

Paper in Journal: New Algorithms and Software for Significance Controlled Variable Selection

VLMCX: Variable Length Markov Chain with Exogenous Covariates

Models categorical time series through a Markov Chain when a) covariates are predictors for transitioning into the next state/symbol and b) when the dependence in the past states has variable length. The probability of transitioning to the next state in the Markov Chain is defined by a multinomial regression whose parameters depend on the past states of the chain and, moreover, the number of states in the past needed to predict the next state also depends on the observed states themselves. See Zambom, Kim, and Garcia (2022) .

CRAN: VLMCX

Package source: SignifReg_3.0.tar.gz

Paper in Journal: soon...