Fault Diagnosis of Batch Reactor Using Machine Learning Methods

Table 6

Results for training of fault features.

Classifiers	Bayes			MLP			RBF
	5 features	9 features	15 features	5 features	9 features	15 features	5 features	9 features	15 features
TP rate	0.542	0.542	0.583	0.917	1	1	1	1	1
FP rate	0.111	0.111	0.092	0.017	0	0	0	0	0
Precision	0.399	0.399	0.384	0.927	1	1	1	1	1
Recall	0.542	0.542	0.583	0.917	1	1	1	1	1
𝐹-measure	0.426	0.426	0.45	0.917	1	1	1	1	1
ROC area	0.815	0.815	0.865	0.989	1	1	1	1	1