Early Detection and Personalized Management of Chronic Obstructive Pulmonary Disease Using Random Forest and SVM Algorithms

Chronic obstructive pulmonary disease (COPD) is a prevalent lung ailment globally and ranks as the third leading cause of mortality. While there is currently no known cure for COPD, timely detection and personalized management can effectively alleviate symptoms and slow down disease progression. In this paper, we propose a machine learning-based approach that utilizes two algorithms, namely Random Forest (RF) and Support Vector Machine (SVM), to identify the severity stages of COPD. The classification of COPD severity stages into four categories, namely Mild, Moderate, Severe, and Very Severe, is based on the GOLD (Global Initiative for Chronic Obstructive Lung Disease) system. Our study utilizes a dataset obtained from the copdgene research study, comprising 5747 samples with 35 attributes. We trained and tested the model using various features to evaluate its accuracy in predicting the severity stages of COPD. The results demonstrate that the trained models achieved high accuracy, with the Support Vector Machine algorithm attaining 84% accuracy and the Random Forest algorithm achieving 87% accuracy. These findings highlight the potential efficacy of our approach in early identification and personalized treatment of COPD, ultimately enhancing patient quality of life and reducing healthcare expenditures.