Evaluating the Risk of Type 2 Diabetes Mellitus Using Artificial Neural Network

Abstract

To identify risk factors, neural network analysis is used to create disease prediction models, including diabetes. The goals of this study were to identify diabetes risk factors and determine their relative contribution using artificial intelligence as a mode of prediction. The current investigation was led by breaking down the dataset, as shown below. We chose a dataset from Kaggle. The diabetes dataset was from India. It has 763 female members, 497 of whom have no diabetes and 266 who have type 2 diabetes. We used neural network analysis to create mathematical models and visualize the distribution of diabetic risk factors. The significance level was set at 0.05. The current study found that the following risk factors were ranked in order of importance: Diabetes Pedigree Function, age, glucose, skin thickness, blood pressure, BMI, insulin, and number of pregnancies. When combined, neural network analysis is effective in developing mathematical models that can predict disease risk factors.