Predicting life expectancy of lung cancer patients after thoracic surgery using SMOTE and machine learning approaches

Abstract

. Lung cancer is a life-threatening condition characterized by the uncontrolled growth and spread of abnormal cells in the lungs. Thoracic surgery is a commonly employed diagnostic and treatment procedure for lung cancer. The objective of this study is to utilize machine learning techniques to predict the life expectancy of lung cancer patients one year after thoraric surgery. The study utilizes the Thoraric Surgery Data Set, consisting of 454 data, with 385 data representing surviving patients and 69 data representing patients who passed away. Due to an imbalance in the data, the Synthetic Minority Oversampling Technique (SMOTE) process is applied to balance the dataset. Multiple machine learning algorithms, including Random Forest (RF), K-Nearest Neighbor (KNN), and Support Vector Machine (SVM), are employed for prediction. Validation is performed using 5-fold cross validation, repeated three times. The results indicate that the KNN model achieves the highest mean accuracy of 84.80% before the SMOTE process, although all models exhibit a low mean F1-score. Following the SMOTE process, the RF model attains the highest mean accuracy of 79.52%, while the KNN model demonstrates the highest mean F1-score of 26.54%. This research contributes valuable insights to clinicians in making informed decisions and improving patient outcomes.