Enhancing Type 2 Diabetes Prediction: A Hybrid Approach with RF & LightGBM

• Author(s): Karthikeyan S ; Nallasivam P V ; Keerthanan G B I ; A. Raihana, M. E.
• Paper ID: 1707757
• Page: 207-212
• Published Date: 08-04-2025
• Published In: Iconic Research And Engineering Journals
• Publisher: IRE Journals
• e-ISSN: 2456-8880
• Volume/Issue: Volume 8 Issue 10 April-2025

Abstract

One major global health concern is diabetes mellitus, especially type 2 diabetes mellitus (T2DM). For T2DM to be effectively managed, early and precise prognosis is crucial. In order to improve accuracy and guarantee adaptability across a variety of datasets, this work presents a prediction model that combines the Random Forest (RF) and LightGBM algorithms. To maximize performance, the model uses extensive preprocessing methods, such as feature selection and data manipulation. The Pima Indians Diabetes Dataset and other publicly accessible datasets were used to assess the suggested methodology, which produced reliable and efficient prediction results. These results demonstrate how RF and LightGBM can be used to create dependable and flexible models for the analysis and prediction of diabetes.

Keywords

Diabetes Prediction, Machine learning, Random Forest, LightGBM

Citations

IRE Journals:
Karthikeyan S , Nallasivam P V , Keerthanan G B I , A. Raihana, M. E. "Enhancing Type 2 Diabetes Prediction: A Hybrid Approach with RF & LightGBM" Iconic Research And Engineering Journals Volume 8 Issue 10 2025 Page 207-212

IEEE:
Karthikeyan S , Nallasivam P V , Keerthanan G B I , A. Raihana, M. E. "Enhancing Type 2 Diabetes Prediction: A Hybrid Approach with RF & LightGBM" Iconic Research And Engineering Journals, 8(10)