Classification of Skin Cancer Images Using Convolutional Neural Networks

Skin cancer is the most common human malignancy according to American Cancer Society. It primarily diagnosed visually, starting with an initial clinical screening and followed potentially by dermoscopic (related skin) analysis, a biopsy histopathological examination. occurs when errors (mutations) occur in DNA of skin cells. The mutations cause cells grow out control form mass aim this study was try classify images lesions help Convolutional Neural Networks. Deep neural networks show humongous potential for image classification while taking into account large variability exhibited environment. Here, we trained on basis pixel values classified them disease labels. dataset acquired from Open Source Kaggle Repository (Kaggle Dataset) which itself ISIC (International Imaging Collaboration) archive. training performed multiple models accompanied Transfer Learning. highest model accuracy achieved over 86.65%. used publicly available ensure credibility reproducibility aforementioned result.