부산대학교 도서관

Melanoma is one of the most deadly diseases in the world, and if not detected early enough, it can spread to other regions of the body. As a response, the medical industry has seen a significant advancement, with the introduction of automated diagnosis tools that may assist doctors and even laymen alike in determining the type of ailment they are dealing with. In this case, we're presenting a hybridized method for detecting melanoma skin cancer that could be applied to any worrisome lesion. An automated skin lesion classification approach is proposed in this study. A deep learning network that has been pre-trained and fine-tuned is used. The performance is then compared using various transfer learning methods. The performance is assessed using well-known quantitative criteria such as specificity, sensitivity, precision, and accuracy.