Classification Of Skin Cancer By Means Of Transfer Learning Models
DOI:
https://doi.org/10.15282/mekatronika.v3i2.7393Keywords:
Skin Cancer, CNN, Classification, Transfer Learning, Machine LearningAbstract
Skin cancer is a disease of human skin affected with abberrant or damaged cell and that lead to the formation of tumours. Skin cancer can be mainly classified into melanoma and non-melanoma, where melanoma is more deadly if misdiagnosis at the early stage. Traditional way of skin cancer classification required dermatologist to classify the cancer based on CT-scan, MRI or X-ray, which may promote risks of misdiagnosis. Hence deep learning is introduced to carry out the image feature extraction for the classification tasks by using the ISIC dataset. With the aids of InceptionV3 on different machine learning model, the skin cancer classification can be carry out by Artificial Intelligence. As a result of this study, Logistic Regression achieved overall classification accuracy of 78.3%, proven it has the ability to classify skin cancer based on skin lesion images
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2021 Ji Zhe Lee, Anwar P. P. Abdul Majeed
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.