Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
6<br />
Name Description getpenmat() first.derivative()<br />
ridge ridge penalty x<br />
penalreg<br />
correlationbased<br />
penalty<br />
lasso lasso penalty x<br />
adaptive.lasso adaptive lasso x<br />
x<br />
+ a.coefs()<br />
fused.lasso fused lasso x<br />
oscar oscar x<br />
scad scad x<br />
weighted.fusion weighted fusion x<br />
bridge bridge x<br />
enet Elastic net x<br />
genet<br />
icb<br />
licb<br />
ao<br />
Generalized<br />
elastic net<br />
Improved<br />
correlationbased<br />
penalty<br />
L 1 -norm<br />
improved<br />
correlationbased<br />
penalty<br />
Approximated<br />
octagon penalty<br />
Table 2: Implemented penalties of class penalty. Whether they consist of getpenmat()<br />
or of first.derivative() and a.coefs() is tagged with an ‘x’.<br />
for some small c > 0, where 1 {βj ≠0} = 1 if β j ≠ 0 and 1 {βj ≠0} = 0 otherwise. Thus we<br />
might use the getpenmat() function.<br />
The source code of the complete implementation of the lasso penalty is given in the<br />
following:<br />
R> lasso