Intelligent Diagnosis of Diabetes based on Information Gain and Deep Neural Network

There are a great deal of diabetes patients all over the world, which has a severe effect on humans. It can cause a variety of serious complications. Therefore, early detection of diabetes is the key to reducing diabetes mortality. In this paper, an IG-DNN algorithm is proposed, which combines Information Gain (IG) and Deep Neural Network (DNN). In the proposed new method the information gain is applied to decrease the attributes to five ones, which are input into DNN network. The new method presented in this paper has a classification accuracy of 90.26%, which is better than most previous research results.