Diabetes Prediction by Artificial Neural Network

Diabetes is a syndrome caused by the hyperglycemia of multiple chronic combined with the variation of carbohydrate, fat and protein metabolism, which impact the improper discharge of insulin and the proper usage of insulin in the human body or both. Diabetes affects more than 463 million people globally. By 2020, 88 million people in Southeast Asia are suffering from this illness. According to a report by the International Federation (IDF), India has 77 million people out of 88 million affected individuals. Therefore, diabetes is one of the growing health concerns in India and has no persistent heal. Therefore, rapid diabetic perception is essential, and it can be done inexpensively through the computation method. The research is carried out for the detection of diabetes by artificial neural network (ANN). Here, the prediction is based on back propagation algorithm of an ANN model for diabetes analysis. For training and testing, the dataset was obtained from the UCI machine learning repository’s Pima Indian Diabetes Dataset (PIDD). The network was built with different neurons at various epochs and observed that the accuracy reaches up to 99.23%. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.