tuneLago {lago} | R Documentation |
Uses a validation set to select tuning parameters for Lago. Based on options implemented so far,
one can use either avgp
or topn
as the performance metric.
tuneLago(Y,X,Ynew,Xnew,alphaGrid,Kgrid,func=avgp,kernel='t',version='s',plot=TRUE,...)
Y |
vector of binary responses; training data. |
X |
matrix of covariates/predictors; training data. |
Ynew |
vector of binary responses; validation data. |
Xnew |
matrix of covariates/predictors; validation data. |
alphaGrid |
a vector listing candidates for the tuning parameter, alpha . |
Kgrid |
a vector listing candidates for the tuning parameter, K . |
func |
function used for performance evaluation; default is func=avgp . |
kernel |
either ‘t’ or ‘g’, specifying the type of kernel to be used by LAGO. |
version |
either ‘s’ or ‘e’, specifying whether sLAGO or eLAGO is to be fitted. |
plot |
if TRUE , a contour plot is generated; see aKcontour . |
... |
arguments passed to the criterion function, func . |
Alexandra Laflamme-Sanders and Mu Zhu, University of Waterloo, Canada.