Using transfer learning and the architecture of Densenet 201 model, deep learning algorithms classification and detection of different skin lesions including skin cancer.
-Original images were augmented to increase the dataset values
-Network training categorical accuracy is 99% and validation categorical accuracy of 81%
-Couldn't upload weights file because of its large size
DATA can be downloaded here : https://www.kaggle.com/kmader/skin-cancer-mnist-ham10000?rvi=1