Lung cancer is a potentially fatal malignancy that is challenging to identify. Because it typically results in death for both men and women, it is especially important that nodules be properly and promptly examined. As a result, a number of methods have been used to identify lung cancer early on. In this approach, many machine learning-based methods for lung cancer diagnosis have been compared. Too many techniques for diagnosing lung cancer have been developed recently, the majority of which rely on CT scan pictures. Additionally, lung cancer nodules are identified using image recognition by combining threshold segmentation algorithms with several classifier techniques. According to this study, CT scan pictures are better suited for reliable outcomes. As a result, CT scan images are primarily utilized to identify malignancy. Additionally, compared to other segmentation methods, marker-controlled threshold segmentation yields more accurate results. Furthermore, the outcomes derived from deep learning approaches were more accurate than those derived from standard machine learning techniques. The accuracy, precision, recall, and F-score of the deep learning algorithm are used to forecast the outcome.