AN EFFICIENT METHOD FOR DIABETES PREDICTION USING LINEAR DISCRIMINANT ANALYSIS AND EXTREME GRADIENT BOOSTING

Diabetes is one of the most common chronic diseases worldwide, and early diagnosis is crucial to prevent complications. In this paper, we propose a method for diabetes prediction that combines Linear Discriminant Analysis (LDA) for dimensionality reduction with the Extreme Gradient Boosting (XGBoost) classifier. Our approach aims to improve the prediction accuracy and robustness of diabetes detection. We evaluate our model using the PIMA Indians Diabetes dataset and demonstrate that our method achieves improved accuracy and robustness. The proposed system is efficient, interpretable, and suitable for real-world healthcare applications. The findings show that our proposed method performed well, achieving better accuracy, precision, and AUC values, demonstrating its effectiveness for diabetes prediction.

کلیدواژه ها:

Diabetes prediction ، Dimensionality reduction ، Extreme Gradient Boosting ، Linear Discriminant Analysis