Detecting High-Risk Factors and Early Diagnosis of Diabetes Using Machine Learning Methods

Table 1

Summary of related work.

S. No.	Ref.	Dataset	Preprocessing method(s)	Outperformed method(s)	Model accuracy (%)
1	[8]	Private	PCA, mRMR	RF	80.84
		PIDD			77.21
2	[13]	PIDD	PCA	SVM, AB, bootstrap	94.44
3	[4]	BRFSS-2014	SMOTE	NN	82.41
4	[14]	BRFSS	Different parameters used	RF	86.80
5	[15]	—	—	LR	77.9
6	[16]	PIDD	Feature selection	NN	86.6
7	[9]	PIDD	Label encoding, normalization	SVM	80.26
		Other		DT, RF	96.81
8	[17]	PIDD	Features extraction	RF	88.31
9	[2]	Private	—	LR	96.02
10	[18]	Private	—	Bagging	97.7