sym.glm: Lasso, Ridge and and Elastic Net Linear regression model to interval variables

Description

Execute Lasso, Ridge and and Elastic Net Linear regression model to interval variables.

Usage

sym.glm(sym.data, response = 1, method = c("cm", "crm"), 
               alpha = 1, nfolds = 10, grouped = TRUE)

Arguments

sym.data

Should be a symbolic data table read with the function read.sym.table(...).

response

The number of the column where is the response variable in the interval data table.

method

"cm" to generalized Center Method and "crm" to generalized Center and Range Method.

alpha

alpha=1 is the lasso penalty, and alpha=0 the ridge penalty. 0

nfolds

Number of folds - default is 10. Although nfolds can be as large as the sample size (leave-one-out CV), it is not recommended for large datasets. Smallest value allowable is nfolds=3

grouped

This is an experimental argument, with default TRUE, and can be ignored by most users.

Value

An object of class "cv.glmnet" is returned, which is a list with the ingredients of the cross-validation fit.

References

Rodriguez O. (2013). A generalization of Centre and Range method for fitting a linear regression model to symbolic interval data using Ridge Regression, Lasso and Elastic Net methods. The IFCS2013 conference of the International Federation of Classification Societies, Tilburg University Holland.

Examples

Run this code

data(int_prost_train)
data(int_prost_test)
res.cm.lasso<-sym.glm(sym.data=int_prost_train,response=9,method='cm',
                      alpha=1,nfolds=10,grouped=TRUE)
pred.cm.lasso<-predictsym.glm(res.cm.lasso,response=9,int_prost_test,method='cm')
plot(res.cm.lasso)
plot(res.cm.lasso$glmnet.fit, "norm", label=TRUE)
plot(res.cm.lasso$glmnet.fit, "lambda", label=TRUE)
RMSE.L(sym.var(int_prost_test,9),pred.cm.lasso)
RMSE.U(sym.var(int_prost_test,9),pred.cm.lasso)
R2.L(sym.var(int_prost_test,9),pred.cm.lasso)
R2.U(sym.var(int_prost_test,9),pred.cm.lasso)
deter.coefficient(sym.var(int_prost_test,9),pred.cm.lasso)