Respiratory disease is common throughout the world, and even during the Covid-19 pandemic, lung ailments were at their peak worldwide. A lot of research has been done and is currently in progress in this field. Some researchers use only “machine learning (ML) or deep learning (DL) techniques”, while others use ML classification techniques with feature selection. This paper focuses on developing a hyper-parameterized hybrid model based on the best machine learning classification techniques for lung disease prediction. The big data framework of healthcare and accurate prediction of respiratory disease will help medical practitioners make decisions and cure diseases. In this article, we propose a new framework, a “hybrid machine learning model based on feature selection, feature optimization and classification techniques” named hybrid machine learning framework (HMLF), which will improve upon existing models in lung disease prediction and overcome the problems of underfitting and overfitting.