DERM-AI: A Convolutional Neural Network Based Approach For Skin Disease Detection

This paper shows how convolutional neural networks (CNNs) are useful for accurately and quickly diagnosing skin issues, underscoring the need of utilizing state-of-the-art machine learning approaches for accurate and fast diagnosis. A CNN architecture including convolutional and pooling layers, fully linked layers, and a softmax classifier was developed to handle the problem of diagnosing skin diseases. The CNN model was trained using stochastic gradient descent, and its hyperparameters were adjusted using grid search and random search. Future research in this area may benefit from the insights the study offers on CNN model construction and application for skin disease diagnosis.