Exploratory Data Analysis and Predictive Modeling of Pima Indian Diabetes

Diabetes, characterized by elevated blood glucose levels, is the world's eighth greatest cause of mortality. In 2012, the World Health Organization recorded over 1.5 million diabetes-related fatalities. Environmental and lifestyle factors have a substantial impact on diabetes prevalence, with a 48% increase projected between 2017 and 2045. Diabetes is incurable, but it can be managed with medicine and lifestyle adjustments. Early detection is crucial for optimal care and the avoidance of serious consequences. This study compares four machine learning algorithms—Random Forest, K-Nearest Neighbors, Decision Tree, and Support Vector Machine—for identifying people at risk of acquiring diabetes. The Random Forest algorithm performed better than the others in terms of prediction accuracy. As a result, a Streamlit-based interface with Random Forest was created to improve the accessibility and effectiveness of diabetes diagnosis and management, supporting prompt interventions and better health outcomes. This approach highlights machine learning's promise for early disease diagnosis and public health management.