An Analysis of the Performance of Machine Learning Algorithms for Prediction of Lung Cancer

Cancer accounted for nearly 10 million deaths in 2020. Prostate, lung, breast, colon, colorectal, stomach, skin, etc are the most common cancers. Lung cancer is the leading cause of deaths related to cancer accounting for estimated 1.8 million deaths in 2020 according to the International Agency for Research on Cancer (IARC). Factors like smoking, occupational hazards, air pollution, chronic lung diseases of the past, and hereditary cancer syndromes can be causes of lung cancer disease. It is often diagnosed at later stages and in the last few decades, Lung Cancer (stage 4) is one of the critical and incurable diseases. Hence beforehand detection and diagnosis is necessary. To detect the cancerous condition various machine learning algorithms have been used in this process due to its accuracy. Logistic regression, naive Bayes, MLP, XGBoost, Decision Tree, random forest, ridge model, SVC, gradient boosting, etc. are some of the algorithms. Their ability to detect lung cancer is well-known. The paper factors that cause lung cancer and application of ML algorithms will be discussed. Two datasets from kaggle have been taken and compared with different algorithms. The data has been preprocessed to improve its reliability. To analyze the performance of an algorithm, accuracy,precision,recall value is used according to different datasets. The data is divided into training and testing in a 7:3 ratio for fair results. People will be able to see pros and cons of different algorithms through this research paper.