GBlockBoost: Computation of the GBlockBoost Algorithm or Componentwise...
In lqa: Penalized Likelihood Inference for GLMs

Description Usage Arguments Details Value Author(s) References See Also

This function fits a GLM based on penalized likelihood inference by the GBlockBoost algorithm. However, it is primarily intended for internal use. You can access it via the argument setting method = "GBlockBoost" in lqa, cv.lqa or plot.lqa. If you use componentwise = TRUE then componentwise boosting will be applied.

1
2
3

   GBlockBoost (x, y, family = NULL, penalty = NULL, intercept = 
       TRUE, weights = rep (1, nobs), control = lqa.control (), 
       componentwise, ...)

`x`	matrix of standardized regressors. This matrix does not need to include a first column of ones when a GLM with intercept is to be fitted.
`y`	vector of observed response values.
`family`	a description of the error distribution and link function to be used in the model. This can be a character string naming a family function, a family function or the result of a call to a family function. See `family()` for further details.
`penalty`	a description of the penalty to be used in the fitting procedure, e.g. `penalty = lasso (lambda = 1.7)`.
`intercept`	a logical object indicating whether the model should include an intercept (this is recommended) or not. The default value is `intercept = TRUE`.
`weights`	some additional weights for the observations.
`control`	a list of parameters for controlling the fitting process. See `lqa.control`.
`componentwise`	if `TRUE` then componentwise boosting will be applied, e.g. there is just a single regressors updated during each iteration. Otherwise GBlockBoost will be applied. If this argument is missing and your penalty is `ridge` then it is set to `componentwise = TRUE`. Note that this will coincide with an application of the RidgeBoost algorithm. In all other cases of a missing argument `componentwise = FALSE` is used.
`...`	further arguments.

The GBlockBoost algorithm has been introduced in Ulbricht \& Tutz (2008). For a more detailed technical description, also for componentwise boosting, see Ulbricht (2010).

GBlockBoost returns a list containing the following elements:

`coefficients`	the vector of standardized estimated coefficients.
`beta.mat`	matrix containing the estimated coefficients from all iterations (rowwise).
`m.stop`	the number of iterations until AIC reaches its minimum.
`stop.at`	the number of iterations until convergence.
`aic.vec`	vector of AIC criterion through all iterations.
`bic.vec`	vector of BIC criterion through all iterations.
`converged`	a logical variable. This will be `TRUE` if the algorithm has indeed converged.
`min.aic`	minimum value of AIC criterion.
`min.bic`	minimum value of BIC criterion.
`tr.H`	the trace of the hat matrix.
`tr.Hatmat`	vector of hat matrix traces through all iterations.
`dev.m`	vector of deviances through all iterations.

Jan Ulbricht

Ulbricht, Jan (2010) Variable Selection in Generalized Linear Models. Ph.D. Thesis. LMU Munich.

Ulbricht, J. \& G. Tutz (2008) Boosting correlation based penalization in generalized linear models. In Shalabh \& C. Heumann (Eds.) Recent Advances in Linear Models and Related Areas. Heidelberg: Springer.

lqa, ForwardBoost

lqa documentation built on May 30, 2017, 3:41 a.m.