Melanoma Skin Cancer Classification using Machine Learning: A Systematic Literature Review

Abstract

One of the deadliest types of skin cancer in the world is Melanoma. The importance of early diagnosis can increase early to improve life. Melanoma is a type of skin cancer that is considered the deadliest type of skin cancer in the world because of its high fatality rate. Early diagnosis of melanoma could help increase the survival rate, making it essential for early identification. However, melanoma diagnosis is subjective and complex, making it difficult to be detected. Many machine learning algorithms were developed to help with melanoma diagnosis automation in the last decade. Those algorithms include K-Nearest Neighbor, Decision Tree, Logistic Regression, Artificial Neural Network, Support Vector Machine, and Deep Learning. This paper aims to summarize six widely used algorithms to inform the readers of how and why they are used and the accuracy of melanoma classification using said algorithms.