You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
REFERENCES 22<br />
control<br />
... further arguments<br />
list of control parameters as returned by <strong>lqa</strong>.control(). See the<br />
<strong>lqa</strong> manual (Ulbricht, 2010a) for details.<br />
This function plots the coefficient build-ups for a given dimension of your tuning parameter(s).<br />
The argument lambdaseq can be omitted. In this case a default sequence<br />
R> lambdaseq plot.<strong>lqa</strong> (y, x, family = binomial (), penalty.family = fused.lasso,<br />
+ offset.values = c (NA, 0.01), add.MLE = FALSE)<br />
The corresponding plot is given in Figure 1. Note that the grouping effect does not<br />
cover x 3 . This is due <strong>to</strong> the small weight λ 2 = 0.01 on the second penalty term that is<br />
responsible for the fusion of correlated regressors. Note that the grouping effect among<br />
x 1 and x 2 is kept all the time. Furthermore, the relevance of the regressors is recognized<br />
as x 4 is selected <strong>to</strong> join the active set at last.<br />
References<br />
Bondell, H. D. and B. J. Reich (2008). Simultaneous regression shrinkage, variable selection<br />
and clustering of predic<strong>to</strong>rs with oscar. Biometrics 64, 115–123.<br />
Chambers, J. (1998). Programming with Data. A <strong>Guide</strong> <strong>to</strong> the S Language. New York:<br />
Springer.<br />
Fahrmeir, L., T. Kneib, and S. Lang (2009). Regression - Modelle, Methoden und Anwendungen<br />
(2nd ed.). Berlin: Springer.<br />
Fahrmeir, L. and G. Tutz (2001). Multivariate Statistical Modelling based on Generalized<br />
Linear Models (2nd ed.). New York: Springer.<br />
Frank, I. E. and J. H. Friedman (1993). A statistical view of some chemometrics regression<br />
<strong>to</strong>ols (with discussion). Technometrics 35, 109–148.<br />
Hastie, T., R. Tibshirani, and J. H. Friedman (2009). The Elements of Statistical Learning<br />
(2nd ed.). New York: Springer.