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