Table 1

Performance of various classification models built for modelling diabetes and hypertension

Type of classificationN for case/controlClassification accuracy at the best random classifier (%)Classification accuracy for the different models used (%)
(i) Diabetes in general population2853/777973.±1.378.6±0.8578.30
(ii) Diabetes in hypertensive population1322/138251.170.987.4±1.175.6±2.772.1
(iii) Two-stage aggregate of (i) + (ii) − diabetes2853/777973.2N/A84.988.2N/A
(iv) Hypertension in general population6759/387363.682.482.4±0.680.0±0.880.9
(v) Hypertension in diabetic population2427/599471.280.180.8±1.376.0±1.467.3
(vi) Two-stage aggregate of (iv) + (v) − hypertension1322/138251.1N/A95.390.3N/A
Kuwait-specific data sets
(i) Diabetes in general population1334/417975.879.479.477.675.9
(ii) Hypertension in general population3451/206262.68079.976.877.9
Asian-specific data sets
(i) Diabetes in general population976/206167.984.384.381.483.6
(ii) Hypertension in general population1933/110463.786.886.883.383.8
  • LR, logistic regression; SVM, support vector machine; k-NN, k-nearest neighbours; MDR, Multifactor Dimensionality Reduction.