After COVID 19, a huge prevalence of pulmonary diseases and their associated complications in lungs which highlights the urgent need for accurate and efficient diagnostic tools to assess severity of lung disease and frame required guidance in treatment strategies. This study was aimed for analyzing the performance of various regression algorithms for the enhancement of diagnostic capabilities for lung disease. A dataset of 38,538 patient records from open-source data repository was utilized for the purpose of evaluating the effectiveness of different regression models based on patient attributes such as symptoms, disease categories, gender and medical history. For the appraisal of performance analysis of the constructed models to detect the severity of lung disease, key performance metrics like Mean absolute error (MAE), Mean Standard Error (MSE) and R2 score were systematically analyzed. In this study the researchers aimed to optimize the best performing regression algorithms after a performance analysis of Random Forest with other multiple regression algorithms for detecting lung disease severity. By systematically varying decision tree parameters of Random Forest Regression and leveraging advanced optimization techniques, cross validation, this research seeks to identify the optimal Random Forest configuration that maximizes key performance metrics such as MAE, MSE, and R2 score. The findings present a comparative analysis of regression methods, affirming the effectiveness of optimized approaches in enhancing model accuracy and reliability. The analysis and optimization procedure contribute the diagnosis and personalized treatment strategies for medical experts.