Analysis and Classification of Skin Cancer Based on Deep Learning Approach

Abstract

Skin cancer has become more dangerous in recent years due to its rapid and widespread spread around the world. This evidence has sparked people’s interest and efforts to develop automatic diagnostic computer systems that can help in early diagnosis. Several approaches based on machine learning and deep learning have been developed to assist in the diagnosis of skin lesions. Our objective in this paper is to conduct a comparative study between different deep learning approaches for skin cancer classification and analysis. Four deep learning-based architectures were studied; ResNet, VGG16, GoogleNet and AlexNet to classify skin cancer into melanoma or non-melanoma. As a finding from our comparative study, the ResNet architecture provided better classification accuracy with a very promising result especially for the False-positives error rate.