Automated Classification of Skin Fungal Infections: A Comparative Study of Machine Learning Techniques

Skin fungal infections are a prevalent dermatological concern, requiring accurate and timely diagnosis for effective treatment. This study develops an automated classification system for skin fungal infections using machine learning techniques. We utilized a dataset of 1,158 images across 8 classes, applying preprocessing techniques such as noise removal, contrast enhancement, and skin tone adjustment to improve image quality. Feature extraction involved deriving color, texture, structural, and statistical features from the preprocessed images. These features were then flattened into feature vectors for classification. Three classifiers were evaluated: Support Vector Machine (SVM), K-Nearest Neighbors (KNN), and Random Forest. The SVM classifier achieved the highest accuracy (74 %), followed by KNN (72.1 %) and Random Forest (67.8%). However, statistical analysis revealed no significant performance differences between the models (p-values > 0.05). These findings suggest that all three classifiers are viable options for automated skin fungal infection classification, with selection depending on factors like computational efficiency or interpretability. This study demon-strates the potential of machine learning in dermatological diagnosis and highlights the importance of preprocessing, feature extraction, and classifier choice in developing automated disease classification systems.