Keywords

Blood Glucose Monitoring, Linear Regression, Non-Invasive Monitoring, Box-Cox Transformation, Diabetes Management, Predictive Modeling, Bonferroni Confidence Intervals

Abstract

Accurate monitoring of blood glucose levels is vital for the management of diabetes, a chronic condition affecting millions worldwide. This study explores a linear regression approach to estimate glucose levels non-invasively using a dataset enriched with demographic, physiological, and sensor-based variables. Following rigorous data preparation, including normalization and encoding, a Box-Cox transformation was applied to address violations of regression assumptions, stabilizing variance and improving model validity. Stepwise selection and hypothesis testing were employed to refne the model, retaining signifcant predictors such as AGE, GENDER, HEARTRATE, and DIABETIC, while excluding variables like NIR Reading and LAST EATEN for their minimal contribution. The fnal model demonstrated robust performance with an adjusted R2 and minimized MSE, underscoring the effcacy of the selected predictors. Notably, AGE and DIABETIC emerged as critical factors infuencing glucose levels. Bonferroni-adjusted confdence intervals validated the signifcance of retained variables, ensuring reliability in prediction. This work highlights the potential of regression-based models for non-invasive glucose monitoring, paving the way for cost-effective and accessible solutions in diabetes care.

Included in

Data Science Commons

Share

COinS