CNN-Based Model for Skin Diseases Classification

In our daily lives, deep learning (DL) is becoming more and more important. Cancer detection, predictive medicine, autonomous vehicles, weather forecasting, and speech recognition are just some of the commercial uses of AI, it has already made a big impact. As pattern recognition algorithms, classifiers need well designed feature extractors since the performance, overall, depends on the quality of training. This study demonstrates the classification of skin diseases using deep learning and convolutional neural networks with deep layers. However, the proposed model was successful in terms of classification accuracy more than (85.8%) thanks to the increase in layer count, improved feature extraction, and precise selection of kernel sizes for each layer. The suggested approach can divide skin conditions into six groups. To the best of our knowledge, the proposed model outperforms state-of-the-art models; moreover, a significate comparison is presented in this research paper.