A Smart Healthcare Recommendation System for Multidisciplinary Diabetes Patients with Data Fusion Based on Deep Ensemble Learning

Table 5

Comparison results of the proposed model with other classifiers before data fusion on the Hospital Frankfurt Germany diabetes dataset.

Classifier model	Acc (%)	Pre (%)	Rec (%)	FM (%)	RMSE	MAE
Logistic regression	77.75	0.71	0.58	0.64	0.22	0.47
Naïve Bayes	76.50	0.67	0.61	0.64	0.24	0.48
Random forest	81.25	0.75	0.69	0.71	0.19	0.43
K-nearest neighbors	77.75	0.71	0.59	0.65	0.22	0.47
Decision tree	83.75	0.73	0.83	0.78	0.16	0.40
Support vector machine	84.00	0.79	0.73	0.76	0.16	0.40
Proposed model	91.00	0.89	0.84	0.86	0.09	0.30