Scopus Indexed Publications

Paper Details


Title
Deciphering Diabetes: Illuminating Prognostic Insights with an Interpretable Machine Learning Framework
Author
, Pritom Saha,
Email
Abstract

Diabetes is a chronic disease that impacts millions of people globally. Early detection and management of diabetes can help mitigate potential complications and enhance patient outcomes. Machine learning (ML) algorithms have become more prevalent for predicting diabetes in recent years. The research paper proposed a machine learning-based approach for predicting diabetes utilizing the Diabetes Health indicator dataset and compared the comprehensive performance of multiple ML algorithms. The study used the LIME (Local Interpretable Model-Agnostic Explanations) package to provide interpretable predictions and acquire insights into the factors contributing to the model’s predictions. Although the ML models were interpretable, they also delivered the highest accuracy 86% demonstrated by the XGBoost classifier in prediction. The LIME produces local explanations for individual predictions, emphasizing the most significant features contributing to the prediction. The study found that, for the XGB model, BMI has a negative influence, while Genhlth positively impacts the outcome. Moreover, Age and Income have a positive impact on the outcome of diabetes. The findings of this research can potentially improve patient outcomes and alleviate the global burden of diabetes.

Keywords
Journal or Conference Name
Smart Innovation, Systems and Technologies
Publication Year
2025
Indexing
scopus