Diabetes Diagnosis and Prediction using Data Mining and Machine Learning Techniques

Diabetes mellitus is a chronic metabolic disorder affecting millions of individuals worldwide, posing a significant healthcare challenge. Early diagnosis and accurate diabetes prediction can improve patient outcomes and reduce healthcare costs. This research uses three machine learning algorithms (K-Nearest Neighbors, Random Forest, and Logistic Regression) to predict diabetes, a significant healthcare problem. A dataset containing 100,000 patients was used, and the K-fold cross-validation method was used to select the best model from the three algorithms based on f1 value, accuracy, precision, and recall. The results showed that the K-Nearest Neighbors algorithm with parameter K=10 produced the highest accuracy of 87% at 10-fold. The Random Forest algorithm produces the highest accuracy of 90.70% at 10-fold. Meanwhile, logistic regression produced the highest accuracy of 88.64% at a 10-fold rate. This research provides an overview of how artificial intelligence algorithms can be applied to classify diabetes status to improve this disease’s management. It provides a basis for effective and accurate decision-making in treating diabetic patients.