Table 6 Comparative study of several related studies for diabetes classification using PIDD.
From: A novel RFE-GRU model for diabetes classification using PIMA Indian dataset
Studies | Model | Accuracy |
|---|---|---|
Çalişir and Doğantekin16 | LDA – MWSVM | 89.74% |
Dadgar and Kaardaan17] | Neural network and genetic algorithm | 87.46% |
Chen et al.18 | DT and K-means | 90.00% |
Haritha et al.19 | Firefly and cuckoo models | 81.00% |
Zhang et al.20 | Feedforward neural network | 82.00% |
Benbelkacem and Atmani21 | RF | 77.00% |
Khanwalkar and Soni22 | Sequential minimal optimization (SMO) | 77.34% |
Maniruzzaman et al.9 | LR | 77.06% |
Patra and Kuntia23 | SDKNN | 83.76% |
Bhoi et al.24 | LR | 76.80% |
Neural networks | 75.80% | |
RF | 75.40% | |
Ramesh et al.25 | LR | 73.30% |
KNN | 79.80% | |
NB | 73.10% | |
SVM + Radial basis function (RBF) | 83.20% | |
Salem et al.26 | DT | 81.89% |
NB | 81.89% | |
Fuzzy-KNN | 90.55% | |
TFKNN | 90.63% | |
Proposed RFE-GRU | Recursive feature elimination and gate recurrent unit | 90.70% |
