Prediction of Covid and Pneunomia Using Ensemble Based Machine Learning Approach

Abstract

Since the release of the novel Covid-19, numerous research projects have been launched worldwide in an effort to accurately forecast it. Since many Covid-19 patients passed away from severe chest congestion, the previous lung illness pneumonia is intimately linked to the latter (pneumonic condition). Even for medical professionals, it might be difficult to distinguish between pneumonia and Covid-19 lung illnesses. The most accurate technique for predicting lung disease is chest X-ray imaging. In this proposed work, the dataset are collected from kaggle. The primary goal is to combine several classifiers to get superior performance over each classifier working alone. In this work local binary pattern are used as the feature extractor and then  combines two distinct machine learning models—random forest and KNN classifiers—before conducting training and testing to provide the desired outcomes using stacking which gives the highest accuracy of 82%.