Fig. 1: Feature distribution of our dataset.

The feature distribution of our dataset (n = 11460) using t-distributed Stochastic Neighbor Embedding (2 components, 1000 iterations). The features are colored according to their type (melanoma or non-melanoma).