English

Subspace exploration: Bounds on Projected Frequency Estimation

Data Structures and Algorithms 2021-01-20 v1 Computational Complexity

Abstract

Given an n×dn \times d dimensional dataset AA, a projection query specifies a subset C[d]C \subseteq [d] of columns which yields a new n×Cn \times |C| array. We study the space complexity of computing data analysis functions over such subspaces, including heavy hitters and norms, when the subspaces are revealed only after observing the data. We show that this important class of problems is typically hard: for many problems, we show 2Ω(d)2^{\Omega(d)} lower bounds. However, we present upper bounds which demonstrate space dependency better than 2d2^d. That is, for c,c(0,1)c,c' \in (0,1) and a parameter N=2dN=2^d an NcN^c-approximation can be obtained in space min(Nc,n)\min(N^{c'},n), showing that it is possible to improve on the na\"{i}ve approach of keeping information for all 2d2^d subsets of dd columns. Our results are based on careful constructions of instances using coding theory and novel combinatorial reductions that exhibit such space-approximation tradeoffs.

Keywords

Cite

@article{arxiv.2101.07546,
  title  = {Subspace exploration: Bounds on Projected Frequency Estimation},
  author = {Graham Cormode and Charlie Dickens and David P. Woodruff},
  journal= {arXiv preprint arXiv:2101.07546},
  year   = {2021}
}
R2 v1 2026-06-23T22:18:33.787Z