Refining Lung Cancer Diagnostics Through Machine Learning Algorithms

Lung cancer remains one of the leading causes of cancer-related mortality worldwide, with increasing incidence and death rates. Early detection significantly improves prognosis but is hindered by the complexity of clinical and molecular data. This study introduces a hybrid machine learning framework that integrates Logistic Regression (LR) and Random Forest (RF) to enhance survivability prediction and subtype classification in lung cancer patients. A dataset comprising 420 patient records was analysed. LR demonstrated outstanding performance with a prediction accuracy of 96%, highlighting its robustness and clinical applicability. In parallel, RF achieved an 89% accuracy rate and effectively identified the most influential risk factors, offering interpretability and insights into variable significance. Comparative analysis with other popular classifiers, including Support Vector Machine (SVM), AdaBoost, K-Nearest Neighbours (KNN), and Extreme Gradient Boosting (XG Boost), revealed LR’s superiority in terms of both accuracy and interpretability. RF’s contribution to feature importance analysis further supports its role in understanding disease progression. The proposed hybrid approach balances predictive performance and transparency, aligning well with the requirements of clinical decision-making systems. These findings underscore the potential of machine learning in supporting early diagnosis and personalized treatment planning in lung cancer care.