An Ensemble based Machine Learning model for Diabetic Retinopathy Classification

As technology and digitization grows, there is a huge surge in digital storage of health records. Machine learning has an important role in uncovering patterns existing in these health records providing interesting insights to medical practitioners for assistance in the diagnosis of various ailments. Due to the sensitivity of the health records, the machine learning algorithms often fail to predict the diseases accurately. In present work, an ensemble based machine learning model comprising of the Machine Learning (ML) Algorithms namely Random Forest classifier, Decision Tree Classifier, Adaboost Classifier, K-Nearest Neighbour classifier, Logistic Regression classifier is experimented on diabetic retinopathy dataset. As a first step, normalization is done on the diabetic retinopathy dataset by min-max normalization method. This normalized dataset is then trained the proposed ensemble model. The performance of the proposed model is finally evaluated against the individual machine learning algorithms. The comparative analysis reveals the fact that the ensemble machine learning model outperforms the individual machine learning algorithms.