Predicting Lung Cancer Using Ensemble Machine Learning Algorithms: A Comprehensive Study

Lung cancer is the leading cause of death from cancers. As such, it demands an earlier and more accurate diagnosis. The current study aims to unearth the different algorithm's performance that's applied for machine learning (ML) for predicting lung cancer-Logistic Regression (LR), Decision Trees (DT), K-Nearest Neighbors (KNN), Naive Bayes, Support Vector Classifier (SVC), Random Forest (RF), Gradient Boosting (GB), and Extreme Gradient Boosting (XGBoost). A clinical feature and symptoms dataset of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{3 0 9}$</tex> samples with <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{1 6}$</tex> columns is used for lung cancer prediction. Here, a Voting Classifier has been employed as an ensemble approach wherein a few models are summed up to acquire one output. XGBoost provided a maximum accuracy of 96.64 % on this dataset. In the case of the Voting Classifier, an overall accuracy of 95.80 % was obtained with a cross-validation of 94.13 %. It also depicts how techniques of ensemble learning are generally very efficient in enhancing the precision and reliability of predictions over lung cancer.