A Hybrid Deep Learning Model for Classification of Psoriasis Disease

Skin disease prediction is a crucial aspect of dermatological diagnosis, often requiring extensive knowledge and expertise. In this work, we propose a hybrid deep learning model based on the efficient net+LSTM architecture for skin disease prediction. The model is trained and tested on a dataset comprising seven classes of psoriasis: Erythrodermic, pustular, plaque, nails, inversus, healthy, and guttate.We compare the performance of three hybrid models: VGG16+LSTM, AlexNet+LSTM, and the proposed efficient net+LSTM. Results show that the efficient net+LSTM model achieves the highest accuracy of 86 percent, surpassing the performance of VGG16+LSTM (66 percent) and AlexNet+LSTM (68 percent).