ampling of a population. Figure 3.19 shows this comparison.
model and a GDA model have also been constructed for the seeds
ble 3.5 shows the confusion matrices of two models. Both had
erformance as LDA. Figure 3.20 shows ROC curves for LR and
dels for the seeds data set. The performance of two models show
significant difference.
The confusion matrices of the LR model and GDA models for the seeds data.
LR
GDA
A
B
%
A
B
%
A
63
8
88.7
61
10
85.9
B
4
61
93.8
3
61
93.8
%
94.0
88.4
90.0
93.8
85.9
88.5
3.20. The ROC curves for the LR and GDA models for the seeds data.
inear algorithms
on to the linear discriminant analysis algorithms, there are several
algorithms for discriminant analysis as well, such as the multi-
ceptron (or artificial neural network), the basis function neural
the support vector machine and the random forest algorithm as
he deep neural network.
