GAMBoost (version 1.2-3)

GLMBoost: Generalized linear model by likelihood based boosting

Description

GLMBoost a convenience wrapper around GAMBoost, for fitting generalized linear models by likelihood based boosting.

Usage

GLMBoost(x,y,penalty=length(y),standardize=TRUE,...)

Arguments

x

n * q matrix of covariates with linear influence.

y

response vector of length n.

penalty

penalty value (scalar or vector of length q) for update of individual linear components in each boosting step. If this is set to 0 the covariates enter the model as mandatory covariates, which are updated together with the intercept term in each step.

standardize

logical value indicating whether linear covariates should be standardized for estimation.

...

arguments that should be passed to GAMBoost

Value

Object returned by call to GAMBoost (see documentation there), with additional class GLMBoost.

References

Tutz, G. and Binder, H. (2007) Boosting ridge regression. Computational Statistics \& Data Analysis, 51(12), 6044--6059.

See Also

GAMBoost, predict.GLMBoost.

Examples

##  Generate some data 
x <- matrix(runif(100*8,min=-1,max=1),100,8)             
eta <- -0.5 + 2*x[,1] + 4*x[,3]
y <- rbinom(100,1,binomial()$linkinv(eta))

##  Fit a model with only linear components
gb1 <- GLMBoost(x,y,penalty=100,stepno=100,trace=TRUE,family=binomial()) 

#   Inspect the AIC for a minimum
plot(gb1$AIC) 

#   print the selected covariates, i.e., covariates with non-zero estimates
getGAMBoostSelected(gb1)

##  Make the first two covariates mandatory

gb2 <- GLMBoost(x,y,penalty=c(0,0,rep(100,ncol(x)-2)),
                stepno=100,family=binomial(),trace=TRUE)