Publication

Abstract : Globally, Diabetes, a chronic metabolic disease having Elevated Blood sugar levels, has serious health consequences. In order to deliver individualized treatment, diabetes management requires accurate blood sugar prediction from predictive models. In this work, we examine how Multilayer Perceptron neural networks can be applied to predict blood sugar levels from blood pressure readings and insulin dosages. The Multi-Layer Perceptron (MLP) neural network design is a potent tool for making predictions about the future. MLPs are neural network architectures that consist of numerous layers of interconnected neurons, which comprise output and hidden layers. The model can recognize complex patterns and relationships because every neuron in one layer is connected to every other layer above it. A dataset comprising Blood Pressure Levels, Blood Sugar Levels, and Insulin levels, Age, BMI from individuals with diabetes has been collected. To minimize prediction errors and optimize network parameters, the backpropagation algorithm is applied.A predictive model's performance is evaluated using a Number of Measures, Sensitivity, and Mean squared Error.