Machine Learning Based Diabetes Detection Model for False Negative Reduction

Diabetes is a chronic disease characterized by the inability of pancreas to produce enough insulin or body’s use efficiently. This becoming increasingly prevalent worldwide and can result in severe complications such as blindness, kidney failure, stroke. Early detection diabetes potentially save millions lives globally, making it crucial focus research. In this study, we propose machine learning model aid predicting diabetes. The comprises several methods: Linear Regression (LnR), Logistic (LR), k-nearest neighbor (KNN), Naive Bayes (NB), Random Forest (RF), Support Vector Machine (SVM), Decision Tree (DT). Prior feeding pre-processed data into for evaluation, conducted pre-processing steps, removing null values, standardizing using normalization, labeling label encoding process. Imbalanced datasets adversely affect accuracy algorithms, address problem balancing Synthetic Minority Oversampling Technique (SMOTE) method. We assessed model’s performance on two found that random forest algorithm produced optimal results, with 97% dataset 2019 80% Pima Indian dataset. However, balanced dataset, significantly reduce number false-negative detections.