NeurIPS 2019
Sun Dec 8th through Sat the 14th, 2019 at Vancouver Convention Center
Paper ID:6721
Title:Selective Sampling-based Scalable Sparse Subspace Clustering


		
This paper describes a variant of sparse subspace clustering in which scalability issues are addressed by running subspace clustering on subsets of the data at a time. Previous work addressed this for random subsets, but this paper considers what amounts to a more targeted sampling approach that leads to non-trivial empirical performance improvements on real datasets and accompanying theory.