Among all types of cancer (breast cancer, bladder cancer, melanoma, leukaemia, lymphoma, and many more), lung cancer is the most prevalent one. The aim of this study is to systematically predict the likelihood of a random person having lung cancer based on some information about them. Five supervised learning model algorithms were used and evaluated for their predictive accuracy. The result of this comprehensive analysis shows that logistic regression, decision tree, random forest, and gradient boost all have the same model accuracy to 93.54% while K-NN is 90.32%. The best model to predict lung cancer in this study is concluded to be a decision tree which has the highest accuracy and receivers operating characteristics (ROC) value of 0.96. Researchers seeking to explore this field in the future should consider expanding the model by conducting an in-depth analysis of the variables used in the models and integrating additional factors to increase forecast accuracy.