Publication

Abstract : Lung cancer, a complex and formidable disease, frequently necessitates surgical intervention as a pivotal aspect of its treatment. Accurate prediction of post-operative life expectancy holds paramount significance in the realm of clinical decisionmaking and patient counselling This study undertook a rigorous comparison of various machine learning algorithms, encompassing Support Vector Classifier (SVC), Logistic Regression, XGBoost, Random Forest, K-Nearest Neighbors (K-NN), and CatBoost. Using an expansive dataset containing clinical and demographic variables collected from lung cancer patients, the research team rigorously assessed the predictive capabilities of these models. The outcome of the research team analysis unveiled exceptional mean accuracy scores across the models, underscoring their utility in this context. Specifically, SVC, Random Forest, and CatBoost consistently exhibited remarkable mean accuracy rates, each reaching 0.864. Logistic Regression also demonstrated robust performance, achieving a commendable mean accuracy of 0.862, while XGBoost and K-NN delivered competitive mean accuracies of 0.846 and 0.84, respectively. These findings underscore the substantial potential of machine learning in furnishing precise predictions regarding post-operative life expectancy. By doing so, these models empower healthcare professionals to make well-informed treatment decisions, paving the path for the development of personalized treatment strategies and the enhancement of patient care in the management of lung cancer.