|
| No | Title | Reference | Advantages | Outcomes |
|
| 1 | “Performance Comparison of Machine Learning Techniques on Diabetes Disease Detection” | [4] | Found that the LR has the best accuracy because of categorical data | DT 75.3%
LR 77.9%%
SVM 77.6%%
KNN 76% |
|
| 2 | “Prediction and Diagnosis of Diabetes Mellitus: A Machine Learning Approach” | [5] | The results show that adaptive boosting with the decision stump's base as a classifier is more accurate | Adaboost%
DT-Base 77.6%%
SVM-Base 77.6%%
DS-Base 80.72% |
|
| 3 | “Prediction of Gestational Diabetes by Machine Learning Algorithms” | [6] | They proved that the ensemble learning used XGBoost has the greatest accuracy | Adaboost 76.2%
GBM 76.5%
XGBoost 77.5% |
|
| 4 | “Analysis and Prediction of Diabetes Diseases using Machine Learning Algorithm: Ensemble Approach” | [7] | It is found that different datasets affect machine learning algorithms' accuracy and single algorithm is less accurate than ensemble learning | Each algorithm has different accuracy on different datasets |
|
| 5 | “Diabetes Prediction Using Different Machine Learning Approaches” | [8] | It has been shown that the SVM method will gain more accuracy when we have no prior knowledge of the data | DT 74%
SVM 82%%
NB 80%%
ANN 81% |
|
| 6 | “The Mahalanobis distance” | [9] | They show the effect of data variance | The MD's results are fewer and more accurate than ED |
|
| 7 | “Machine Learning Prediction Models for Gestational Diabetes Mellitus: Meta-analysis” | [2] | The meta-analysis and findings of heterogeneity were done with the help of the Meta Disc software | Age, heredity, BMI, and fasting blood glucose were the most common features used to build models |
|
| 8 | “A comparative analysis of KNN, GA, SVM, DT, and LSTM algorithms in machine learning” | [10] | The performance of five essential machine learning algorithms is compared and demonstrated | SVM algorithm has provided one of the best results in predictive analytics in real-time applications |
|
| 9 | “Discovering Tree Based Diabetes Prediction Model” | [11] | This study focuses on essential features; this puts a lot of effort into the data mining and reduces the complexity of predicting model | Feature selection, and prediction model by DT algorithm, obtain a good result |
|
