Table 1.

Summary of machine learning techniques for diabetes detection.

Sr. No.	Reference	Year	Methodology	Finding and Results
1	Sisodia et al. [12]	2018	Naive Bayes, SVM, and DT	NB classifier outperformed the other classifiers with an accuracy of 76.30%.
2	Naz et al. [13]	2020	Artificial Neural Network (ANN), Bayes, Decision Tree, and Deep Learning	Deep Learning (DL) attained the highest 98.07% accuracy
3	Khanam et al. [14]	2021	SVM, DT, k-Nearest Neighbours (kNN), Random Forest (RF), Logistic Regression (LR), AdaBoost (AB), and Neural Network (NN)	Neural Network (NN) outperformed the other techniques and reached an accuracy of 88.6% on the Pima Indians Diabetes (PID) dataset
4	Hasan et al. [15]	2020	Weighted ensemble model of kNN, DT, RF, AB, NB, and XGBoost	The results demonstrated that the proposed ensemble classifier achieved 78.9% sensitivity, 93.4% specificity, and 95% AUC
5	Singh et al. [16]	2021	Ensemble model of DT, RF, SVM, XGBoost, and NN	The model demonstrated an accuracy of 95%.
6	Pradhan et al. [17]	2020	Artificial neural network	With 70% training data and 30% testing data, the model achieved an accuracy of 85.09%
7	Kannadasan et al. [18]	2019	Deep Neural Network (DNN)	The results demonstrated that the classifier achieved an accuracy of 86.26%.
8	Maniruzzaman et al. [22]	2017	Linear Discriminant Analysis (LDA), Quadratic Discriminant Analysis (QDA), and Naive Bayes (NB)	The results demonstrated that the model achieved an accuracy of 81.97%.
9	Azad et al. [19]	2021	Synthetic Minority Over-sampling Technique (SMOTE), Genetic Algorithm (GA), and Decision Tree (DT)	The model was tested on the Pima Indians Diabetes (PID) dataset and achieved an accuracy of 82.1256%.
10	Kumari et al. [20]	2021	k-Nearest Neighbours (kNN)	The model achieved an accuracy of 92.28% on the diabetes dataset
11	Abokhzam et al. [21]	2021	Machine Learning grid-based Random Forest	The model achieved an accuracy of 95.7%.