<![CDATA[Diabetes is one of the major health issues worldwide, affecting 10.5% of the total adult population (20-79 years old). Often referred to as the silent killer, nearly half of those affected by diabetes are unaware of their condition. Diabetes is categorized into several types, namely type 1 diabetes mellitus, type 2 diabetes mellitus, and gestational diabetes. Detection of diabetes can be carried out through various methods, including blood sugar level tests, Hemoglobin A1c (HbA1c) tests, oral glucose tolerance tests, as well as physical examinations and medical history reviews by doctors. Interpreting the results of these tests can be used to identify the potential for an individual to have diabetes, employing a machine learning approach as a decision support system for doctors to make informed decisions, and also providing patients with reminders to consult with a doctor. In the machine learning model we've developed, we trained and tested algorithms using the 'Diabetes prediction dataset,' consisting of 8 variables: age, gender, Body Mass Index (BMI), hypertension, heart disease, smoking history, HbA1c level, and blood glucose level. The algorithm employed was the Artificial Neural Network (ANN) with the optimizer using Stochastic Gradient Descent (SGD). This application is intended to serve as a decision support system for doctors and the general public. It's designed using Anvil for 8 types of input variables, providing 2 output variables: the percentage of an individual's potential to have diabetes and suggestions for preventing such risks.]]>
Copyrights © 2024
