Small Sample Spaces Cannot Fool Low Degree Polynomials

  • Authors:
  • Noga Alon;Ido Ben-Eliezer;Michael Krivelevich

  • Affiliations:
  • School of Mathematics, Institute for Advanced Study, Princeton, NJ 08540, USA, and Schools of Mathematics and Computer Science, Raymond and Beverly Sackler Faculty of Exact Sciences, Tel Aviv Univ ...;School of Computer Science, Raymond and Beverly Sackler Faculty of Exact Sciences, Tel Aviv University, Tel Aviv, Israel 69978;School of Mathematical Sciences, Raymond and Beverly Sackler Faculty of Exact Sciences, Tel Aviv University, Tel Aviv, Israel 69978

  • Venue:
  • APPROX '08 / RANDOM '08 Proceedings of the 11th international workshop, APPROX 2008, and 12th international workshop, RANDOM 2008 on Approximation, Randomization and Combinatorial Optimization: Algorithms and Techniques
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

A distribution Don a set $S \subset \mathbb Z_{p}^{N}$ 茂戮驴-fools polynomials of degree at most din Nvariables over $\mathbb Z_{p}$ if for any such polynomial P, the distribution of P(x) when xis chosen according to Ddiffers from the distribution when xis chosen uniformly by at most 茂戮驴in the 茂戮驴1norm. Distributions of this type generalize the notion of 茂戮驴-biased spaces and have been studied in several recent papers. We establish tight bounds on the minimum possible size of the support Sof such a distribution, showing that any such Ssatisfies $$ |S|\geq c_{1} \cdot \left(\frac{(\frac{N}{2d})^{d} \cdot \log{p}}{\epsilon^{2}\log{(\frac{1}{\epsilon})}}+p\right). $$This is nearly optimal as there is such an Sof size at most $$ c_{2} \cdot \frac{(\frac{3N}{d})^{d} \cdot \log{p} + p}{\epsilon^{2}}. $$