Volume 18, No. 6, 2021
Enhancing Lung Cancer Detection: A Comparative Efficiency Study Of Machine Learning Supervised And Unsupervised Models
M. Sheik Mansoor and M. Mohamed Sathik
Abstract
Lung cancer is one of the most aggressive types of cancer, known for its rapid spread. It metastasizes through lymphatic fluid and the bloodstream, reaching organs like the bones, glands, and brain. The incidence of lung cancer is rising significantly due to air pollution and industrial contaminants. The World Health Organization (WHO) predicts that lung cancer-related deaths could reach 9.6 million by 2020, highlighting the severity of the issue. Early detection of lung cancer is crucial for effective treatment, but despite the availability of manual CT scan analysis in the medical field, it remains challenging for doctors to accurately determine the stage of the disease from these images. Consequently, to predict lung cancer and its varieties early on, the medical informatics research community has developed a number of machine learning models. This research compares two well-known supervised learning models, Artificial Neural Networks (ANN) and Support Vector Machines (SVM), using data on lung cancer that was taken from the Cancer Image Archive. We also applied the dataset to two unsupervised learning models, Apriori and K-means, to examine the differences in performance. The final results, including performance metrics such as accuracy, precision, and recall, were compared and displayed in a table.
Pages: 9981-9990
Keywords: Machine Learning; Lung Cancer Prediction; Supervised Learning; Cancer Diagnosis.