Disease Diagnosis and Prediction Framework using an Ensemble Classifier

Abstract

The significance of digitization in healthcare has increased due to the rapid advancement of technology. Individuals with chronic illnesses need to be monitored closely and treated right away. The possibility of major consequences is decreased by early identification of these disorders. Algorithms for machine learning are frequently employed to forecast this kind of illness. In order to forecast diabetics, we provide a novel ensemble-based machine learning technique in this work. A range of machine learning methods, such as Naïve Bayes, Support Vector Machine, Decision Tree, k-NN, and Random Forest, were employed to analyse 16769 records in total. With a 97% accuracy rate, the best ensemble based SNDT classifier is selected through voting method.  Recall, accuracy, and F1-score are among the other performance metrics that are assessed.