FWDselect (version 1.1)

FWDselect-package: FWDselect: Selecting variables in regression models

Description

FWDselect package introduces a simple method to select the best model using different types of data (continuous, binary or poisson) and applying it in different contexts (parametric or nonparametric). The proposed method is a new forward stepwise-based selection procedure that selects a model containing a subset of variables according to an optimal criterion (obtained by cross-validation) and also takes into account the computational cost. Additionally, bootstrap resampling techniques are used to implement tests capable of detecting whether significant effects of the unselected variables are present in the model.

Details

ll{ Package: FWDselect Type: Package Version: 1.1 Date: 2013-02-01 License: GPL } FWDselect is just a shortcut for ``Forward selection'' and is a very good summary of one of the package's major functionalities, i.e., that of providing a forward stepwise-based selection procedure. This software helps the user select relevant variables and evaluate how many of these need to be included in a regression model. In addition, it enables both numerical and graphical outputs to be displayed.

The package includes several functions that enable users to select the variables to be included in linear, generalized linear or generalized additive regression models. Users can obtain the best combinations of q variables by means of the main function selection. Additionally, if one wants to obtain the results for more than one size of subset, it is possible to apply the qselection function, which returns a summary table showing the different subsets, selected variables and information criterion values. The object obtained when using this last function is the argument required for plot.qselection, which provides a graphical output. Finally, to determine the number of variables that should be introduced into the model, only the test function needs to be applied.

References

Burnham, K., Anderson, D. (2002). Model selection and multimodel inference: a practical information-theoretic approach. 2nd Edition Springer.

Efron, B. (1979). Bootstrap methods: another look at the jackknife. Annals of Statistics, 7, 1--26.

Efron, B. and Tibshirani, R. J. (1993). An introduction to the Bootstrap. Chapman and Hall, London.

Miller, A. (2002). Subset selection in regression. Champman and Hall.

Sestelo, M., Villanueva, N. M. and Roca-Pardinas, J. (2013). FWDselect: an R package for selecting variables in regression models. Discussion Papers in Statistics and Operation Research, 13/02.