Lung Cancer Prediction using Machine Learning

Abstract

Prevention is always a better option than cure, especially when it comes to deadly disease like cancer. According to the latest reports by the International Agency for Research on Cancer, Lung cancer is the second leading cause of death. Along with the family history, main causes for lung cancer are Genetic Factors, Smoking and the unhealthy lifestyle in the world. In a developing country like India, cancer treatments are very costly and hard to access to all the sections of the society. In a country where 220 million Indians sustained with an expenditure level of Rs 32/day, affording cancer treatments is impossible. Therefore, predicting the disease acts as the saviour to the millions of people in the country. For this purpose, we have identified the specific genes responsible for causing lung cancer in the Human race. Selecting a small number of genes can lead to a better accuracy. In this paper , we have used Kruskal-Wallis test. This helped us to select the genes expression data. Finally, we have identified12 influential genes responsible for causing lung cancer. The accuracy of the model is 84.375% using the Random Forest algorithm. All the files and codes used in the work is available at https://github.com/Farazkhan0516/Lung-Cancer-Prediction-using-Machine-Learning.git.