atures stands for the number of independent variables, and

used to specify the graphic style,

vip(model,num_features,geom)

e 4.24 shows how this package was used for comparing the

ce measurements for the above three constrained linear

n models constructed for the olive oil content data. The result is

ifferent from the output of the glmnet function, which always

he fruit weight as the most important positive contributor to the

content. This is because a vip model is constructed using the

idation approach. It shows that three models generated a

cy result. The stone weight, stone width and fruit weight

were the most important variables. In addition, the paste water

d as a consistent negative contributor to the oil content in olives.

(a) (b) (c)

The vip visualisation of the importance of the constrained regression models

for the olive oil content data. (a) The RLR model. (b) The Lasso model. (c)

model.

nonlinear regression analysis algorithms

onlinear algorithms have been introduced in the last chapter for

tion analysis, such as neural networks, support vector machine

om forest. They can also be used for regression analysis. This