Lung Cancer Classification Using Machine Learning

Lung cancer is leading among nineteen different types of cancer deaths across the world. Later the lung cancer is detected, the lesser the chances of survival for an individual. The authors have proposed a model to predict lung cancer in the initial stages. To build the model, the authors applied seven different classifiers: Naive Bayes (NB), K Nearest Neighbors (KNN), Decision Tree (DT), Random Forest (RF), Adaboost, Logistic Regression (LR), and Xgboost, to the lung cancer dataset accessible on Kaggle. The authors further applied class balancing using the SMOTE technique and observed the effectiveness of classifiers. The authors found that the class balancing improved the performance. LR has achieved the highest accuracy (96.30%) among all applied classifiers after applying class balancing.