Abstract—Data mining is the computational process to find potentially useful knowledge from the stored information and then use the discovered knowledge to predict or classify the new data item that its target class was unknown. Among many available algorithms to do data classification or prediction, Bayesian network (BN) is one of the most accurate methods. BN is acyclic directed graph that models probabilistic dependencies among variables with a conditional probability distribution. In this research, we propose prediction model of the postoperative survival in the lung cancer patients using BN. Currently, cancer is one of the leading causes of morbidity and mortality worldwide. The top causes of cancer death is lung cancer. Lung cancer surgery is one of treatment methods, but this method is risky. Sometimes patients died after the surgery. We thus study the surgery risk using BN. We show the performance of the proposed BN technique through the specific application for predicting post-operative life expectancy in the lung cancer patients from the Wroclaw Thoracic Surgery Centre, Poland. The experimental results show that the BN with appropriate discretization and learning scheme can predict the one year survival after surgery as accurate as 91.28%.
Index Terms—Data prediction, Bayesian network, Lung cancer, discretization, tree-augmented Naïve Bayes.
K. Sriwong is with the School of Computer Engineering, Suranaree University of Technology (SUT), 111 University Avenue, Muang, Nakhon Ratchasima 30000, Thailand (e-mail: firstname.lastname@example.org).
K. Kerdprasop is with the School of Computer Engineering and Knowledge Engineering Research Unit, SUT, Thailand (e-mail: email@example.com).
N. Kerdprasop is with the School of Computer Engineering and Data Engineering Research Unit, SUT, Thailand (e-mail: firstname.lastname@example.org).
Cite: Kittipat Sriwong, Kittisak Kerdprasop, and Nittaya Kerdprasop, "Post-Operative Life Expectancy of Lung Cancer Patients Predicted by Bayesian Network Model," International Journal of Machine Learning and Computing vol. 8, no. 3, pp. 280-285, 2018.