S.No

Author(year)

Method(Accuracy)

1

Sarwar et. (2018)

KNN(77), SVM(77), LR(74), RF(71)

2

Sonar and Prof.K.JayaMalini (2019)

DT(85), NB(77), SVM(77.3)

3

Soni et al. (2020)

RF(77)

4

Daanouni et al. (2020)

ANN(87.5), DT(82.50)

5

L.J.Muhammad et al. (2020)

GB(88.76), RF(88.76)

6

Xu and Wang (2019)

RF(93), XG(93)

Column attributes

Description

Diabetes

Yes (60), No (330)

Waist/Hip

Ratio may be a more significant heart

disease risk factor than BMI.

Waist/Hip

In inches

Diastolic BP

The lower number of blood pressure

Systolic BP

The upper number of blood pressure

BMI

703 x weight (lbs)/ [height(inches)]

Weight

In pounds

Height

In inches

Gender

162 males, 228 females

Age

All adult African Americans

Chol/HDL

Ratio of total cholesterol to good

cholesterol. The desirable finding is less than 5

HDL

Good Cholesterol

Glucose

Fasting blood sugar

Cholesterol

Total cholesterol

Patient number

Identifies patients by number

Measures

Formula

ROC

Trade-off between True positive rateand False Positive Rate

F1-Score

F=2*(P*R)/(P+R)

Recall(R)

R=TP/(TP+FN)

Precision(P)

P=TP/(TP+FP)

Accuracy(A)

A=TP+TN/(TP+TN+FP+FN)

Algorithm

Withoutparame- ter tuning

Randomizedsearch CV

Gridsearch CV

Decision tree

92.3%

94%

92.3%

Random Forest

94%

93.2%

93.2%

LogisticRe-gression

94.9%

94.9%

94.9%

SVM

94%

93.2%

94.7%

Ada Boost

90%

94%

92.3%

Light GBM

94%

91.4%

92.3%

Gradient Boost

91.5%

90.6%

91.5%

Cat Boost

94.9%

95.7%

94%

Algorithm

PIMAIndiaDataset

Proposed model

Decision Tree

71.2%

94%

Random For-est

77.48%

94%

Logistic Re-gression

74.89%

94.9%

SVM

74.09%

94.9%

Ada Boost

75.32%

94%

Light GBM

75%

94%

GradientBoost

75.75%

91.5%

Cat Boost

75.32%

95.7%