Diabetes Prediction and Recommendations Using Machine Learning and Generative AI
Abhishek Ponnaboina1, D.Veeraswamy2
1Department of Electronics and Communication Engineering Institute of Aeronautical Engineering, Hyderabad, INDIA abhi04457@gmail.com
2Assistant Professor, Department of Electronics and Communication Engineering
Institute of Aeronautical Engineering, Hyderabad, INDIA Veeraswamy44@gmail.com
ABSTRACT
Machine learning along with artificial intelligence (AI) seem to make a paradigm shift into the healthcare domain by ensuring better diagnosis and treatment approaches. Considering the Mexico’s increasing prevalence of dia- betes, timely diagnosis and properly determined interven- tion approach becomes essential. This project presents a unique AI-based web application that simulates and eval- uates diabetes risk in individuals with the help of machine learning and generative AI, enabling personalized health recommendations. Specifically, the application evaluates eight parameters such as glucose, blood pressure, BMI, age, etc. All of the parameters will allow users to foresee the risk of diabetes and determine the appropriate response to it. Moreover, the application uses Google Gemini, a generative AI model, to provide customized diabetes interventions by analyzing a patient’s health data, and determines likely future health scenarios. Such interven- tions relate and explain the causes of diabetes, abnormal characteristics of human health, and effective dietary and lifestyle changes. The system is built using Streamlit which allows for a graphic interface that is convenient whereby the users input health data and instantly get diagnosis and interventions. It can be inferred that there is a huge opportunity in the realm of health technology advancing in terms of better preventive healthcare, which results in better patient outcomes as well as better integration of machine learning and AI.
Keywords—Diabetes prediction, machine learning, generative AI, healthcare analytics, Streamlit, supervised learning.
