Diabetes Mellitus Prediction Using Artificial Neural Network

Student: Joshua Oluwaseun Awolusi (Project, 2025)
Department of Computer and Information Science
Bamidele Olumilua University of Edu. Science and Tech. Ikere Ekiti, Ekiti State

Abstract

Diabetes mellitus is a significant global health concern characterized by chronic high blood sugar levels and associated complications. Early prediction and diagnosis are crucial for effective management and prevention of the disease. This project presents a novel approach to predict diabetes mellitus using Artificial Neural Networks (ANNs), leveraging their ability to model complex nonlinear relationships in medical data. The dataset utilized for this study comprises diverse features, including demographic, clinical, and lifestyle factors. Data preprocessing steps, including normalization and handling of missing values, were implemented to enhance the quality of the input data. In this study, an automatic diabetes prediction system has been developed using the database derived from Mendeleys named “ DATASET 2 - EARLY-STAGE DIABETES RISK PREDICTION DATASET”.The database was split into training data of 80% which was used to train and evaluate the model to access its predictive capabilities and testing data of 20% which was used to test the model which made it 98% accurate. The data was used to develop machine and deep learning classifiers to predict diabetes

Keywords
diabetes mellitus prediction artificial neural network artificial
Filters
Institutions