Skin lesion classification on the HAM10000 dataset using a ConvNeXt-Small model with transfer learning. Includes data preprocessing, augmentation, class imbalance handling, and evaluation across seven diagnostic categories.
tensorflow convolutional-neural-networks confusion-matrix transfer-learning data-augmentation f1-score melanoma skin-cancer ham10000 skin-cancer-detection dermoscopic-images medical-image-classification convnext precision-and-recall convnext-small imagenet-pretraining
-
Updated
Dec 25, 2025 - Jupyter Notebook