Intelligent Diabetes Detection System Using Machine Learning Techniques

Artificial intelligence (AI) techniques are increasingly employed to analyze diabetic growth patterns (DGP) and assess human glucose patterns (HGP) in potential diabetic patients (PDPs). While HGP data alone is often insufficient for accurately predicting potential diabetic outcomes (PODs), incorporating additional demographic, environmental, and clinical features significantly enhances diagnostic accuracy. The Kaggle datasets provide a diverse range of features, which, when combined with dimension reduction techniques such as Principal Component Analysis (PCA), significantly improve the model’s ability to detect relevant patterns and reduce computational complexity. To address these challenges, we propose a novel three-phase system that integrates k-means clustering for data preprocessing, PCA for dimensionality reduction, and a logistic regression model to identify hidden relationships within the data. The model’s performance is evaluated using critical metrics, including F1-score, precision, accuracy, and recall. Our results demonstrate the effectiveness of this approach, achieving 99% accuracy, 99% precision, 99% recall, and 99% F1-score, significantly outperforming several traditional classification models.