Performance Comparison of XGBoost and LightGBM Gradient Boosting Algorithms in Predicting Cervical Cancer Risk

This paper compares two renowned gradient- boosting techniques XGBoost and LightGBM, comprehensively in the prediction of cervical cancer in its setting. Identification of cervical cancer early, is vital for effective treatment due to its health significance. Binary Classification is applied using a comprehensive dataset of Iifecycle, medical, and demographic attributes. To determine the effectiveness of these models, performance indicators such as Fl-score, Recall, Accuracy, binary_log_loss, precision and roc _ auc curve are employed. Additionally, computational efficiency is scrutinized concerning memory usage and execution time. The investigation includes a pivotal feature significance analysis that identifies the essential factors in cervical cancer prediction. The findings contribute valuable insights and also ensure that the applicability of LightGBM and XGBoost is thoroughly assessed. This perspective aids in choosing the best algorithm for addressing the pressing healthcare concern of cervical cancer, where the paramount obj ectives of early detection and precise prediction are emphasized, resulting in improved patient outcomes.