ResearchTrend.AI
  • Papers
  • Communities
  • Events
  • Blog
  • Pricing
Papers
Communities
Social Events
Terms and Conditions
Pricing
Parameter LabParameter LabTwitterGitHubLinkedInBlueskyYoutube

© 2025 ResearchTrend.AI, All rights reserved.

  1. Home
  2. Papers
  3. 1609.00265
27
11
v1v2 (latest)

Testing kkk-Monotonicity

1 September 2016
C. Canonne
Elena Grigorescu
Siyao Guo
Akash Kumar
K. Wimmer
ArXiv (abs)PDFHTML
Abstract

A Boolean kkk-monotone function defined over a finite poset domain D{\cal D}D alternates between the values 000 and 111 at most kkk times on any ascending chain in D{\cal D}D. Therefore, kkk-monotone functions are natural generalizations of the classical monotone functions, which are the 111-monotone functions. Motivated by the recent interest in kkk-monotone functions in the context of circuit complexity and learning theory, and by the central role that monotonicity testing plays in the context of property testing, we initiate a systematic study of kkk-monotone functions, in the property testing model. In this model, the goal is to distinguish functions that are kkk-monotone (or are close to being kkk-monotone) from functions that are far from being kkk-monotone. Our results include the following: - We demonstrate a separation between testing kkk-monotonicity and testing monotonicity, on the hypercube domain {0,1}d\{0,1\}^d{0,1}d, for k≥3k\geq 3k≥3; - We demonstrate a separation between testing and learning on {0,1}d\{0,1\}^d{0,1}d, for k=ω(log⁡d)k=\omega(\log d)k=ω(logd): testing kkk-monotonicity can be performed with 2O(d⋅log⁡d⋅log⁡1/ε)2^{O(\sqrt d \cdot \log d\cdot \log{1/\varepsilon})}2O(d​⋅logd⋅log1/ε) queries, while learning kkk-monotone functions requires 2Ω(k⋅d⋅1/ε)2^{\Omega(k\cdot \sqrt d\cdot{1/\varepsilon})}2Ω(k⋅d​⋅1/ε) queries (Blais et al. (RANDOM 2015)). - We present a tolerant test for functions f ⁣:[n]d→{0,1}f\colon[n]^d\to \{0,1\}f:[n]d→{0,1} with complexity independent of nnn, which makes progress on a problem left open by Berman et al. (STOC 2014). Our techniques exploit the testing-by-learning paradigm, use novel applications of Fourier analysis on the grid [n]d[n]^d[n]d, and draw connections to distribution testing techniques.

View on arXiv
Comments on this paper