Abstract: Malignant skin development is a usual occurrence of cancer. Melanoma, which is also a brutal type of skin cancer, is the most fatal type of epidermis illness, which accounts to 76 percent of deaths caused by skin cancer, even though the occurrence of this cancer is not so common . The most efficient strategy to fight this is to try to figure it out in the earlier stage and medicate the disease with minimal medical procedure. In this study, I specifically focus on skin cancer and make use of more advanced, larger, and greater purpose of CNN which improve the execution. In light of these assumptions, I suggest developing a computerised skin cancer recognisation model based on the analysis of skin damage images using EfficientNet - B6, which records finer details of the cancer. The trial results on the ISIC 2020 Challenge Dataset, which was established by the ISIC and include images from a few key clinical sources, revealed cutting-edge order execution when compared to previous prominent on the equivalent dataset of melanoma classifiers.

Keywords: Malignant Skincancer, Melanoma, CNN, EfficientNet-B6.

| DOI: 10.17148/IARJSET.2022.9693