AI-Powered Prediction of Diabetes for Improved Clinical Decisions

Abstract

Here, there is an intelligent predictor mechanism which identifies the risk of diabetes very early and accurate and which also predicts with precision on the medical decision making and records of the patients. The system can use machine learning procedures done on patient records in Pima Indians Diabetes Database and create outputs by indicating individuals who have a more probable risk of developing diabetes. Thus, it is already similar to the conventional/standard diagnostic procedures, which can reduce delayed effects in the future in the form of morbidity. We use Support Vector Machines (SVM), Random Forest, and Logistic Regression to analyze a range of machine learning algorithms and compare their representation based on numerous criteria, such as precision and recall, accuracy, and F1-score. As a first result, the AI model that we have created provides a high rate of accuracy both in terms of prediction and a larger number of points compared to the other current systems that have been in use until now. Therefore, the instrument will also assist the health care providers to have proactive knowledge to proactive action with regards to intervention time and personal interventions strategies in the event of diabetes at the public health sector.