Lung cancers are identified as one of the lethal diseases by medical professionals due to delays in diagnosis leading to high mortality rates. Early detection of lung cancer improves survival probabilities, but standard diagnosis methods entail high expenses and lengthy examination times with susceptibility to human errors. Thus, this study aims to automate lung cancer prediction using machine learning and deep learning models utilizing a dataset with 16 numerical attributes. GNB, SVM, Logistic Regression, Decision Tree, Random Forest, Gradient Boosting, and XGBoost, and DL models like CNN, MobileNet and Swin Transformer were tested utilizing hyperparameter tuning together with cross validation approaches. The XGBoost model achieved the highest accuracy of 0.9968 during cross-validation tests using the stratified k-fold (k = 5) and leave-one-out methods. XGBoost and Gradient Boosting demonstrated optimal performance after hyperparameter tuning, as they achieved an accuracy of 0.9968 for both training and testing sets, although the total training time was relatively different. CNN demonstrated powerful performance throughout its training and testing stages, achieving the fastest training time among deep learning models with accuracy values of 0.9829 and 0.9872. Ensemble ML methods and optimized DL models were highly effective in lung cancer prediction. Future work will investigate the application of large-scale data platforms to improve the predictive performance of DL models.