Leveraging Machine Learning for Accurate COPD Detection

The driving force behind this research is to use the power of machine learning models to predict the of chronic obstructive pulmonary disease (COPD) in patients by analyzing correlations of various medical factors related chronic obstructive pulmonary disease (COPD). A dataset from Mexican hospital was taken from Kaggle to carry forward this research. COPD is a typical lung disease and implementation of machine learning algorithms, Cat Boost and Logistic Regression to diagnose the disease with the help of dataset which contains relevant features related to COPD. Integrating these algorithms into model this research gives valuable outcomes into the factors which contribute more to the disease aims to increase the accuracy and effectiveness of COPD diagnosis. Further, to give valuable insights into the factors which contribute more to the disease. The research concludes to show that some factors such as Cardiovascular diseases, Age, Hypertension show maximum correlation with presence of COPD, whereas certain conditions such as obesity diabetes show minimal to no correlation. Algorithms, Random Forest and Categorical Boosting showed best performance in terms of accuracy and confusion matrix.