Cancer is still one of the most common causes of death, for men, as well as women. The early detection of cancer is a major contributor that significantly improves treatment outcomes and survival rates. Lung cancer, in particular, poses a substantial risk of misdiagnosis. However, it is the speed of lung cancer diagnosis that often marks the difference between life and death. An early diagnosis of lung cancer has the capability of saving a lot of lives, where a delay in diagnosis can cause the patient to go through hours of surgery, and, in the worst-case scenario, death. The recent development of machine learning suggests that not only does it have the potential to improve the accuracy of lung cancer detection but also the time it takes to diagnose. This Research paper focuses on the evaluation and direct comparison of several machine learning models, such as Neural Networks, Support Vector Machines (SVM), Random Forests, Logistic Regression, and Ensemble Learning Methods, which are used to predict lung cancer. Firstly, the criterion for the performance of the models is obtained through methods such as accuracy, precision, recall, and F1-score, which are used to determine the most optimal predictive model. The results of the research shown in this paper are part of the AI-driven diagnostics domain, and they are based on the strong as well as the weak pieces of the learning model application in lung cancer prediction.