Unsupervised Co-Learning on $G$-Manifolds Across Irreducible Representations

Fighting with thecurse of dimensionalityby leveraging low-dimensional intrinsic structures has become an important guiding principle in modern data science.