DiaDetect: Detecting and Exploring the Effect of Diabetes using Machine Learning
Students & Supervisors
Student Authors
Supervisors
Abstract
The early detection and accurate diagnosis of diabetes are important to lower complications and improve patient outcomes and are one of the fastest-growing global health considerations. The research in this paper offers an improved machine learning model with Support Vector Machine (SVM) technique from patient's medical information to predict diabetes. The method proposed in this work involves complete data processing, imputation of missing values based on the outcomes of that processing, optimization of the class using SMOTE, feature engineering, feature selection and hyperparameter tuning, in order to improve the performance of the prediction process. Several machine learning models were tested and the results gave a better performance using the optimized SVM. In the experimental evaluation, a high accuracy was obtained 90%, precision of 90.58%, a high recall rate at 90%, and an F1 score of 89.96%, which means that the model has good generalization capacity and capability of classification. The results indicates that the proposed model may be a reliable, efficient, and cost-effective method for detecting diabetes which would be suitable for use in real-world diabetes applications in clinical decision support systems in healthcare applications.
Keywords
Publication Details
- DOI: 10.1109/QPAIN69676.2026.11545704
- Type of Publication:
- Conference Name: 2026 IEEE 2nd International Conference on Quantum Photonics, Artificial Intelligence & Networking
- Date of Conference: 16/04/2026 - 16/04/2026
- Venue: IT Business Incubator, Chittagong University of Engineering and Technology (CUET), Chattogram, Bangladesh
- Organizer: IEEE Photonics Society Bangladesh Chapter