Life Expectancy Post Thoracic Surgery Using Machine Learning

The scope of this paper is to propose a life expectancy rate and examine the mortality after thoracic surgery which takes into account the different importance of various features which can have an effect in the end result. The data of the patients collected after diagnosis have been used as the dataset. Various metrics which affect the result have been analyzed with the help of random forest and decision tree algorithms to better understand the consequences of post-surgery. Particular metrics have been selected according to their weightage on the main outcome for prediction. Thus, it enables us to have better comprehension with various algorithms and also some important parameters are selected for the construction of a better model. In addition, we have several classification features such as presence of pain before surgery, hemoptysis before surgery, cough before surgery, whether the patient is a smoker, whether the patient has asthma, and a few others. This classification model predicts whether the patient will survive for a year-long period or not with better selection of the data features.