English

Nonembeddability of Persistence Diagrams with $p>2$ Wasserstein Metric

Functional Analysis 2019-10-31 v1 Machine Learning Algebraic Topology Metric Geometry

Abstract

Persistence diagrams do not admit an inner product structure compatible with any Wasserstein metric. Hence, when applying kernel methods to persistence diagrams, the underlying feature map necessarily causes distortion. We prove persistence diagrams with the p-Wasserstein metric do not admit a coarse embedding into a Hilbert space when p > 2.

Keywords

Cite

@article{arxiv.1910.13935,
  title  = {Nonembeddability of Persistence Diagrams with $p>2$ Wasserstein Metric},
  author = {Alexander Wagner},
  journal= {arXiv preprint arXiv:1910.13935},
  year   = {2019}
}