Rank-Transformed Dissimilarity Profiles for High-Dimensional Classification

Xiangbo Mo, Hao Chen

Despite advances in representation learning, high-dimensional classification remains challenging in low-sample-size regimes, where the dominant signal may vary across applications…