step_fmx {QuantileGH}R Documentation

Forward Selection of ghgh-parsimonious Model with Fixed Number of Components KK

Description

To select the ghgh-parsimonious mixture model, i.e., with some gg and/or hh parameters equal to zero, conditionally on a fixed number of components KK.

Usage

step_fmx(
  object,
  test = c("BIC", "AIC"),
  direction = c("forward", "backward"),
  ...
)

Arguments

object

fmx object

test

character scalar, criterion to be used, either Akaike's information criterion AIC-like, or Bayesian information criterion BIC-like (default).

direction

character scalar, 'forward' (default) or 'backward'

...

additional parameters, currently not in use

Details

The algorithm starts with quantile least Mahalanobis distance estimates of either the full mixture of Tukey gg-&-hh distributions model, or a constrained model (i.e., some gg and/or hh parameters equal to zero according to the user input). Next, each of the non-zero gg and/or hh parameters is tested using the likelihood ratio test. If all tested gg and/or hh parameters are significantly different from zero at the level 0.05 the algorithm is stopped and the initial model is considered ghgh-parsimonious. Otherwise, the gg or hh parameter with the largest p-value is constrained to zero for the next iteration of the algorithm.

The algorithm iterates until only significantly-different-from-zero gg and hh parameters are retained, which corresponds to ghgh-parsimonious Tukey gg-&-hh mixture model.

Value

Function step_fmx returns an object of S3 class 'step_fmx', which is a list of selected models (in reversed order) with attribute(s) 'direction' and 'test'.

See Also

step


[Package QuantileGH version 0.1.7 Index]