Post Thoracic Surgery Life Expectancy Prediction Using Machine Learning

Abstract

Lung cancer survival rate is very limited post-surgery irrespective of if it is small cell or non-small cell. A lot of work has been carried out by employing machine learning in life expectancy prediction post thoracic surgery for patients with lung cancer. Many machine learning models like multi-layer perceptron (MLP), SVM, naïve Bayes, decision tree, random forest, logistic regression have been applied for post thoracic surgery life expectancy prediction based on data sets from UCI. Also, work has been carried out towards attribute ranking and selection in performing better in improving prediction accuracy with machine learning algorithms. So accordingly, the authors, here, have developed a deep neural network-based approach in prediction of post thoracic life expectancy which is the most advanced form of neural networks. This is based on dataset obtained from Wroclaw Thoracic Surgery Centre machine learning repository which contained 470 instances. On comparing the accuracy, the results indicate that the deep neural network can be efficiently used for predicting the life expectancy.