Predictive Modelling for Diabetes Risk Assessment: An Exploratory Data Analysis Approach

Abstract

Diabetes mellitus is a big health problem in the world. More people are getting it, and it can cause bad health issues. This study builds a model to find diabetes early. It uses a machine learning tool to identify the best result. The study uses full exploratory data analysis (EDA). It finds key risk signs like blood glucose level, body mass index (BMI), age, and insulin levels. These things are looked at to see how they are linked and how they affect who gets diabetes. The data’s class imbalance is fixed using resampling. This makes the model more true and fair. It is also easy to see why. It shows the key causes of diabetes. This work helps find diabetes early. It helps doctors make choices and check risk for people with diabetes. It does this by using ideas from data and machine learning tools