Classification of Diabetic Data using Novel Optimized Deep Ensemble Learning

Machine learning application in medical data analysis is one of the most popular approaches in the diagnosis of various diseases. Authors in this work have classified the diabetic data collected from the UCI machine learning repository by using a modified ensemble machine learning approach. In the proposed method, a stacked ensemble model with a different meta-classifier is considered for classification. In the first stage, the performance is compared with the Support vector machine and gradient boosting meta-model. Again the learning rate for the meta model has been changed to get better accuracy. From the result, it is observed that the modified model is providing better results compared to others.