Open AccessImplementasi DenseNet Untuk Mengidentifikasi Kanker Kulit Melanoma
Author(s) -
Jasman Pardede,
Dwi Adi Lenggana Putra
Publication year - 2020
Publication title -
jutisi (jurnal teknik informatika dan sistem informasi)
Language(s) - English
Resource type - Journals
ISSN - 2443-2229
DOI - 10.28932/jutisi.v6i3.2814
Subject(s) - skin cancer , melanoma , convolutional neural network , cancer , basal cell , basal cell carcinoma , human skin , medicine , artificial intelligence , dermatology , computer science , cancer research , pathology , biology , genetics
Skin is a part of a human body that covers the entire body and protect the lower layer from direct sunlight and another microorganism. Because of that, skin cells are always changing and could be changed because of genetic mutation that causes skin cancer. In general, skin cancer is divided into three groups, namely : skin cancer Basal cell carcinoma, skin cancer Squamous cell carcinoma, and skin cancer Melanoma. Melanoma skin cancer is caused by abnormal growth in melanocyte cells. Several methods are proposed to predict Melanoma skin cancer using ResNet, LeNet, and Support Vector Machine. System performance is measured based on the value of accuracy, precision, recall, and f-measure. This experiment is conducted using a Melanoma skin cancer dataset that obtained the average value in terms of accuracy, precision, recall, and f-measure are 0.94, 0.95, 0.92, and 0.94 respectively. Based on that result, the proposed DenseNet121 performs better with 0.94 accuracy, compared with ResNet, LeNet, and Support Vector Machine method.
Keywords— Convolutional Neural Network; Image Classification; Melanoma Classification; DenseNet121.