A Machine Learning based Approach to Early Stage Diabetes Prediction

Diabetes is one of the leading causes of increasing mortality rates. An unhealthy lifestyle, urbanization, obesity/overweight, genetics, hormonal imbalance, poor diet, smoking, and alcoholism have all been linked to an increase in diabetes. Diabetes is highly dangerous if left undiagnosed for an extended period of time, as it can lead to life-threatening complications such as stroke and heart disease. It is possible to develop efficient and effective solutions to detect diabetes in its early stages by applying Machine Learning (ML) algorithms to real-world problems. Several ML models used for early diabetes detection are compared and analyzed in this research paper. SVM, DT, Random Forest, XGBoost, KNN, and Logistic Regression are among the categorization techniques considered in this research study. The hyperparameters of the models are tuned to achieve optimal performance using grid search. The performance of the proposed algorithm is evaluated by using various performance metrics such as precision, accuracy, recall, F1-Score, and ROC-AUC Curve.