Diabetes is one of the most feared diseases currently faced by humanity. The disease is due to a poor reaction of the body to insulin: it is an important hormone in our body that converts sugar into energy that is necessary for the proper functioning of a normal life. Diabetic disease has serious complications on our body because it increases the risk of developing kidney disease, heart disease, retinal disease, nerve damage, and blood vessels. In this article, we have proposed a decision tree model: SLDPS (Diabetes Prediction System with Supervised Learning). The data set is collected via IoT sensors. The classification accuracy obtained with this model was improved to 94.63% after the rebalancing of the data set and shows a potential relative to other classification models in the literature. © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021.