Enhancing Skin Disease Classification: A Hybrid CNN-SVM Model Approach

The multiple skin diseases that are covered in this study are reviewed together with their signs, root causes, and a way of solving a problem. Skin illnesses include many ailments, from common problems like psoriasis, eczema, and acne to uncommon, complicated disorders. This is important to understand the genesis, causative variables, and consequences of these diseases on humans way of life. This method involves using an infected hand to classify different classes, pooling layers, and other elements that make the model. This study included 5494 pictures in total. To assess the model's performance, several metrics such as precision, recall, F1-score, support, accuracy, and average are used. Based on these criteria, the effectiveness of the model is calculated and it has an overall accuracy of 75.25% and an average weighted F1 score of 75.28%. The model attained the highest precision score of 82.77% and the lowest precision score of 65.38%. The model achieved the highest recall of 84.06% and the lowest recall of 69.07%. In the field of skin disease, there is a significant importance for developing efficient early skin disease detection techniques. These tactics not only contribute to better disease management but also provide people with skin conditions with helpful, effective options. This research will help the doctor for treatment as well as patients for prevention.