Comprehensive Analysis of Melanoma Detection using CNN methods

Melanoma is a rare kind of skin cancer that is mostly caused by UV radiation. It appears in the form of pigmented lesions and develops from melanocytes. Much research has been conducted to assist in the classification of melanoma and non-melanoma using machine learning or deep learning models. To get the most accurate outcomes, researchers applied a variety of approaches such as the ABCDE method, SVM, KNN, CNN, Deep CNN, and transfer learning. Despite the fact that such algorithms give great accuracy, there are no medical organizations that rely entirely on them. Another constraint might be the number of melanoma datasets. Because it is such a rarely seen cancer the model will have difficulty categorizing the skin lesions. Previous studies show that researchers have used both machine-learning models and conventional CNN techniques. The accuracy using these was far lower than what is obtained using recent technologies. This study addresses all of these shortcomings and examines previous works/research.