Deep neural network ensemble by data augmentation and bagging for skin lesion classification
This work addresses disease classification in dermatology, but it appears incremental as it builds on existing methods for a specific dataset.
The authors tackled skin lesion classification by developing a deep neural network ensemble using data augmentation and bagging to address data scarcity and overfitting, achieving results for the ISIC 2018 challenge.
This work summarizes our submission for the Task 3: Disease Classification of ISIC 2018 challenge in Skin Lesion Analysis Towards Melanoma Detection. We use a novel deep neural network (DNN) ensemble architecture introduced by us that can effectively classify skin lesions by using data-augmentation and bagging to address paucity of data and prevent over-fitting. The ensemble is composed of two DNN architectures: Inception-v4 and Inception-Resnet-v2. The DNN architectures are combined in to an ensemble by using a $1\times1$ convolution for fusion in a meta-learning layer.