Table 1.

State-of-art techniques used in Type-2 diabetes Prediction.

Technique	Algorithm	Data Type	Overall Accuracy	Remark
ML Technique Ahamed et al. [2]	LR	Tabular Data	75.2 %	Used in assessing diabetes from pre-existing data using feature selection.
	XGB		83.3 %
	GBC		94.1 %
	DT		94.4 %
	RF		94.8 %
	LGBM		95.2 %
ML Technique Kulkarni et al. [38]	XGB	ECG data	96.8 %	The study is limited to the early detection of diabetes.
ML Technique Ahmed et al. [39]	NB	Tabular Data	86.1 %	Prediction of diabetes from the pre-existing tabular data by feature selection.
	DT		96.8 %
	GBC		91.0 %
CA Approach Saha and Saha [40]	RCT	Real-time blood sample data	95.0 %	The approach is invasive and needs frequent finger pricking.
ML Technique Shen et al. [41]	NB	IoT and Embedded systems for real-time data	84.1 %	The approach is invasive and needs finger pricking.
	J48		99.7 %
	LR		86.0 %
	RF		99.6 %