Skin cancer, the most prevalent type of human malignancy, is generally detected visually, beginning with a clinical screening or perhaps followed by dermoscopic analysis, a biopsy, and microscopic testing. Due to the fine-grained variance in the appearance of skin lesions, automated classification of skin lesions using images is a complicated question.
The dataset contains two classes of skin cancer, which are given below:
- Benign
- Malignant
In this I am going to identify two distinct classes of moles using a Convolution Neural Network with Deep Learning Based pre-trained Transfer Learning Model with additional layers for enhancement , and then analyze the results to determine how the model may be used in a real context.
- Pre-processed Images
- Model Performance
- Prediction
- Confusion Matrix