English

Strong Self-Concordance and Sampling

Data Structures and Algorithms 2020-07-13 v2

Abstract

Motivated by the Dikin walk, we develop aspects of an interior-point theory for sampling in high dimension. Specifically, we introduce a symmetric parameter and the notion of strong self-concordance. These properties imply that the corresponding Dikin walk mixes in O~(nνˉ)\tilde{O}(n\bar{\nu}) steps from a warm start in a convex body in Rn\mathbb{R}^{n} using a strongly self-concordant barrier with symmetric self-concordance parameter νˉ\bar{\nu}. For many natural barriers, νˉ\bar{\nu} is roughly bounded by ν\nu, the standard self-concordance parameter. We show that this property and strong self-concordance hold for the Lee-Sidford barrier. As a consequence, we obtain the first walk to mix in O~(n2)\tilde{O}(n^{2}) steps for an arbitrary polytope in Rn\mathbb{R}^{n}. Strong self-concordance for other barriers leads to an interesting (and unexpected) connection -- for the universal and entropic barriers, it is implied by the KLS conjecture.

Keywords

Cite

@article{arxiv.1911.05656,
  title  = {Strong Self-Concordance and Sampling},
  author = {Aditi Laddha and Yin Tat Lee and Santosh Vempala},
  journal= {arXiv preprint arXiv:1911.05656},
  year   = {2020}
}
R2 v1 2026-06-23T12:14:46.482Z