Implementation of AI in Skin Disease Classification Using Convolutional Neural Network Method (CNN)

Abstract

This research aims to develop and test a skin disease classification model using the Convolutional Neural Network (CNN) method by leveraging the MobileNetV2 and VGG16 transfer learning architectures. The dataset used was obtained from Kaggle, which includes various types of skin diseases such as cellulitis, impetigo, athlete's foot, nail fungus, ringworm, cutaneous larva migrans, chickenpox, and shingles. The research process includes data collection and processing, model training, model performance evaluation based on accuracy, and model comparison results. The research results show that MobileNetV2 excels in computational efficiency and training speed, achieving higher accuracy than VGG16 at 96.14%. This research is expected to assist medical professionals in diagnosing skin diseases more quickly and accurately.