Diabetes Prediction using Data Mining Techniques

Abstract

This project delves into the realm of data mining for diabetes prediction, with a focus on enriching the well- known Pima Indian Diabetes dataset by incorporating vital health metrics, particularly heart rate, derived from fitness trackers. The initiative encompasses three integral components: firstly, the exploration and enhancement of the dataset through the integration of heart rate; secondly, the application of diverse data mining techniques, ranging from logistic regression and random forest to support vector machines and naive Bayes, along with advanced machine learning algorithms and neural networks, for a comprehensive comparative analysis of their efficacy in diabetes prediction; and finally, the development of a user-friendly web application disseminating diabetes-related information, detailing project methodologies, and offering users the ability to assess their diabetes risk through instant feedback based on inputted health details. This multi-faceted approach not only contributes to the field of healthcare data analytics by evaluating diverse prediction methods but also aligns with contemporary trends in wearable technology integration for health monitoring, bridging the gap between data analytics and public health awareness. The findings offer valuable insights for healthcare professionals and researchers striving for robust methods in diabetes prediction and prevention.