Research on Predicting Lung Cancer with Clinical Variables

Abstract

At present, research on lung cancer prediction mainly focuses on accurate prediction of lung cancer, and there is little research on predicting the risk of lung cancer based on clinical factors. In this study, a binary logistic regression model was used to process a dataset of lung cancer-related factors from Kaggle, which was uploaded to the website in 2022 and contains 309 samples and 15 influencing factors. Based on a logistic regression of 15 factors, it was found that only Smoking, Yellow fingers, Peer pressure, Chronic Disease, Fatigue, Allergy, Alcohol Consuming, all these 9 factors have a strong correlation with lung cancer. The use of these clinical factors to determine whether there is a risk of lung cancer has appeared less frequently in previous studies, providing partial support for people to prevent lung cancer by changing their lifestyle habits, and pointing out the direction for future in-depth research in this area.