NORMA eResearch @NCI Library

Classification of Melanoma Skin Cancer from Melanocyte Cell Images using Transfer Learning Techniques

Nayak, Soumya (2023) Classification of Melanoma Skin Cancer from Melanocyte Cell Images using Transfer Learning Techniques. Masters thesis, Dublin, National College of Ireland.

[thumbnail of Master of Science]
PDF (Master of Science)
Download (2MB) | Preview
[thumbnail of Configuration manual]
PDF (Configuration manual)
Download (2MB) | Preview


Melanoma skin cancer, one of the most extreme skin diseases plaguing the world and a reason for large fatality rate has been the subject of broad research for many years. In order to identify Melanoma skin cancer, many Deep Learning approaches have given significant predictions. Since all the classes of skin lesion images have almost identical symmetry and optics, it is highly challenging to obtain accurate results and predictions, and this demands for a significant improvement in existing models. With the idea to enhance these results overlying in existing methods, this research focuses on implementing pre-trained transfer learning models for Melanoma skin cancer classification. Three different transfer learning models MobileNetV2, InceptionV3 and DenseNet201 were implemented by fine tuning the models to enhance the overall performance and make true predictions. The performance of all three models were compared and results shows that DenseNet201 outperformed MobileNetV2 and InceptionV3 models by achieving test accuracy and sensitivity of 68.30%.

Item Type: Thesis (Masters)
Milosavljevic, Vladimir
Subjects: Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science
Q Science > Life sciences > Medical sciences > Pathology > Tumors > Cancer
Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions: School of Computing > Master of Science in Data Analytics
Depositing User: Tamara Malone
Date Deposited: 23 May 2023 15:39
Last Modified: 23 May 2023 15:39

Actions (login required)

View Item View Item