Abstract
Diabetes is a chronic disease that causes deaths all around the world, and become worse if not detected at early stage. Therefore, it is essential to detect the disease at early stages before it starts to damage the human body. At this point, detecting diabetes becomes a significant process that need to be performed efficiently. To predict diabetes risk, machine learning algorithms can be used to make prediction when a proper dataset is given. This research introduces some potentially useful machine learning algorithms to predict diabetes risk, guides the practical usage of these algorithms and discusses the comparative results. On this manner, three different machine learning algorithms and one feature selection method are implemented to an open-source dataset. The implementation results presented the applicability of the algorithms to predict diabetes risk.