On the Hardness of the One-Sided Code Sparsifier Problem

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Grigorescu, Elena, Moayyedi, Alice
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914072756224000
author Grigorescu, Elena
Moayyedi, Alice
author_facet Grigorescu, Elena
Moayyedi, Alice
contents The notion of code sparsification was introduced by Khanna, Putterman and Sudan (arxiv.2311.00788), as an analogue to the the more established notion of cut sparsification in graphs and hypergraphs. In particular, for $α\in (0,1)$ an (unweighted) one-sided $α$-sparsifier for a linear code $\mathcal{C} \subseteq \mathbb{F}_2^n$ is a subset $S\subseteq [n]$ such that the weight of each codeword projected onto the coordinates in $S$ is preserved up to an $α$ fraction. Recently, Gharan and Sahami (arxiv.2502.02799) show the existence of one-sided 1/2-sparsifiers of size $n/2+O(\sqrt{kn})$ for any linear code, where $k$ is the dimension of $\mathcal{C}$. In this paper, we consider the computational problem of finding a one-sided 1/2-sparsifier of minimal size, and show that it is NP-hard, via a reduction from the classical nearest codeword problem. We also show hardness of approximation results.
format Preprint
id arxiv_https___arxiv_org_abs_2510_03184
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle On the Hardness of the One-Sided Code Sparsifier Problem
Grigorescu, Elena
Moayyedi, Alice
Information Theory
E.4; F.2.2
The notion of code sparsification was introduced by Khanna, Putterman and Sudan (arxiv.2311.00788), as an analogue to the the more established notion of cut sparsification in graphs and hypergraphs. In particular, for $α\in (0,1)$ an (unweighted) one-sided $α$-sparsifier for a linear code $\mathcal{C} \subseteq \mathbb{F}_2^n$ is a subset $S\subseteq [n]$ such that the weight of each codeword projected onto the coordinates in $S$ is preserved up to an $α$ fraction. Recently, Gharan and Sahami (arxiv.2502.02799) show the existence of one-sided 1/2-sparsifiers of size $n/2+O(\sqrt{kn})$ for any linear code, where $k$ is the dimension of $\mathcal{C}$. In this paper, we consider the computational problem of finding a one-sided 1/2-sparsifier of minimal size, and show that it is NP-hard, via a reduction from the classical nearest codeword problem. We also show hardness of approximation results.
title On the Hardness of the One-Sided Code Sparsifier Problem
topic Information Theory
E.4; F.2.2
url https://arxiv.org/abs/2510.03184