Machine learning classification of polycystic ovary syndrome based on radial pulse wave analysis

Table 4 Performance evaluation of models

Models	Training results (Stratified k-fold cross-validations)			Testing results
Models	Accuracy	AUC	F1 score	Accuracy	AUC	F1 score
KNN	72.672 ± 6.259	0.691 ± 0.105	0.823 ± 0.040	69.565	0.680	0.795
SVM	72.387 ± 4.543	0.696 ± 0.098	0.825 ± 0.027	72.174	0.694	0.818
Decision Trees	72.395 ± 4.307	0.624 ± 0.070	0.819 ± 0.032	71.304	0.646	0.811
Random Forest	72.689 ± 5.188	0.689 ± 0.118	0.825 ± 0.033	69.565	0.695	0.798
Logistic Regression	70.933 ± 3.874	0.698 ± 0.111	0.814 ± 0.028	72.174	0.704	0.814
Voting	73.546 ± 4.232	0.701 ± 0.115	0.831 ± 0.027	72.174	0.715	0.818
LSTM	74.135 ± 5.437	0.702 ± 0.115	0.828 ± 0.023	72.174	0.722	0.818

ISSN: 2662-7671