On the Noise Stability of Small De Morgan Formulas
نویسندگان
چکیده
We show a connection between the De Morgan formula size of a Boolean function f : {0, 1}n → {0, 1} and the noise stability of the function. Specifically, we prove that for 0 < p ≤ 1/2 it holds that NSp(f) ≥ 1− 2p √ L(f) · ‖f‖2 · (1− ‖f‖2) where NSp(f) is the noise stability of f with noise parameter p, ‖f‖ is the L2 norm of f , and L(f) is the De Morgan formula size of f . This result stems from a generalization of Khrapchenko’s bound [Khr71], that might be of independent interest. Our main result implies the following lower bound: ∑ S⊆[n] δ|S|f̂(S)2 ≥ ‖f‖ − 1− δ 2 √ L(f)‖f‖2(1− ‖f‖2) for 0 ≤ δ ≤ 1, where f̂(S) is the Fourier coefficient of f at S. In particular, this bound implies a concentration result on the spectrum of Boolean functions that can be computed by small De Morgan formulas. Specifically, for any ε > 0, we show that ∑ S⊆[n], |S|<k f̂(S)2 ≥ ‖f‖2 (1− ε) where k is roughly 1 2ε √ L(f)1−‖f‖ 2 ‖f‖2 . We observe that this concentration result also stems from a relation between the average sensitivity of f and the original Khrapcheko bound. In addition, we show that the De Morgan formula size in the results mentioned above can be replaced by the square of the non-negative quantum adversary bound, thus giving a (possibly) tighter bound. ∗Weizmann Institute of Science, Rehovot 76100, Israel. Email: {anat.ganor, ilan.komargodski}@weizmann.ac.il. Research supported by an Israel Science Foundation grant and by the I-CORE Program of the Planning and Budgeting Committee and the Israel Science Foundation. †Centre for Quantum Technologies, National University of Singapore, Singapore 117543. Email: [email protected]. ‡Weizmann Institute of Science, Rehovot 76100, Israel and Institute for Advanced Study, Princeton, NJ. Email: [email protected]. Research supported by an Israel Science Foundation grant, by the ICORE Program of the Planning and Budgeting Committee and the Israel Science Foundation, and by NSF grants number CCF-0832797, DMS-0835373.
منابع مشابه
Small bias requires large formulas
A small-biased function is a randomized function whose distribution of truth-tables is small-biased. We demonstrate that known explicit lower bounds on the size of (1) general Boolean formulas, (2) Boolean formulas of fan-in two, (3) de Morgan formulas, as well as (4) correlation lower bounds against small de Morgan formulas apply to small-biased functions. As a consequence, any strongly explic...
متن کامل#SAT Algorithms from Shrinkage
We present a deterministic algorithm that counts the number of satisfying assignments for any de Morgan formula F of size at most n3−16ε in time 2n−n ε · poly(n), for any small constant ε > 0. We do this by derandomizing the randomized algorithm mentioned by Komargodski et al. (FOCS, 2013) and Chen et al. (CCC, 2014). Our result uses the tight “shrinkage in expectation” result of de Morgan form...
متن کاملDually quasi-De Morgan Stone semi-Heyting algebras I. Regularity
This paper is the first of a two part series. In this paper, we first prove that the variety of dually quasi-De Morgan Stone semi-Heyting algebras of level 1 satisfies the strongly blended $lor$-De Morgan law introduced in cite{Sa12}. Then, using this result and the results of cite{Sa12}, we prove our main result which gives an explicit description of simple algebras(=subdirectly irreducibles) ...
متن کاملDually quasi-De Morgan Stone semi-Heyting algebras II. Regularity
This paper is the second of a two part series. In this Part, we prove, using the description of simples obtained in Part I, that the variety $mathbf{RDQDStSH_1}$ of regular dually quasi-De Morgan Stone semi-Heyting algebras of level 1 is the join of the variety generated by the twenty 3-element $mathbf{RDQDStSH_1}$-chains and the variety of dually quasi-De Morgan Boolean semi-Heyting algebras--...
متن کاملImpulsive Noise Elimination Considering the Bit Planes Information of the Image
Impulsive noise is one of the imposed defectives degrades the quality of images. Performance of many image processing applications directly depends on the quality of the input image. Hence, it is necessary to de-noise the degraded images without losing their valuable information such as edges. In this paper we propose a method to remove impulsive noise from color images without damaging the ima...
متن کامل