Prediction of Second Primary Lung Cancer Patient’s Survivability Based on Improved Eigenvector Centrality-Based Feature Selection

Modeling of second primary lung cancer (SPLC) patients' survival prediction has important theoretical significance and practical needs. Cancer survivability may provide advice for better clinical decisions personalized medicine. The Surveillance, Epidemiology, End Results (SEER) program provides large data sets analysis with machine learning methods. SPLC cases are identified labeled from the SEER database; set is then preprocessed improved eigenvector centrality-based feature selection (IECFS). IECFS method utilizes interclass intraclass dispersions ranking criteria. By adjusting value α parameter number features selected, achieves best performance. experiment divided into five folds. This yields a accuracy 90.998% five-year that higher than original classification (89.16%) other state-of-the-art For three-year survivability, proposed methods 83.16%, slightly outperforming all compared effective generalizable.