Risk prediction of hyperuricemia based on particle swarm fusion machine learning solely dependent on routine blood tests

Table 4 Accuracy of each model for 10 folds

Models	Fold1	Fold2	Fold3	Fold4	Fold5	Fold6	Fold7	Fold8	Fold9	Fold10
Lr + smote	0.761	0.757	0.779	0.759	0.739	0.764	0.789	0.776	0.773	0.740
lr + adasyn	0.759	0.759	0.745	0.754	0.750	0.746	0.756	0.764	0.760	0.748
Svm + smote	0.916	0.934	0.918	0.900	0.895	0.928	0.936	0.927	0.909	0.922
Svm + adasyn	0.934	0.924	0.912	0.941	0.922	0.932	0.926	0.934	0.926	0.927
Rf + smote	0.927	0.919	0.923	0.921	0.913	0.923	0.942	0.935	0.934	0.912
Rf + adasyn	0.923	0.940	0.935	0.943	0.934	0.942	0.937	0.930	0.924	0.922
Xgboost + smote	0.962	0.974	0.969	0.964	0.951	0.967	0.969	0.964	0.970	0.962
Xgboost + adasyn	0.969	0.974	0.975	0.974	0.970	0.971	0.969	0.966	0.971	0.975
Dnn + smote	0.901	0.967	0.965	0.956	0.968	0.976	0.968	0.986	0.976	0.981
Dnn + adasyn	0.918	0.965	0.964	0.970	0.968	0.979	0.984	0.986	0.991	0.984

ISSN: 1472-6947