The threshold degree of a Boolean function f is the minimum degree of a real polynomial p that represents f in sign: f(x) = sgn p(x). Introduced in the seminal work of Minsky and Papert (1969), this notion is central to some of the strongest algorithmic and complexity-theoretic results for constant-depth circuits. One problem that has remained open for several decades, with applications to computational learning and communication complexity, is to determine the maximum threshold degree of a polynomial-size constant-depth circuit in n variables. The best lower bound prior to our work was Ω(n(d-1)/(2d-1)) for circuits of depth d. We obtain a polynomial improvement for every depth d, with a lower bound of Ω(n3/7) for depth 3 and Ω(√n) for depth d ≥4. The proof contributes a novel approximation-theoretic technique of independent interest, which exploits asymmetry in circuits to prove their hardness for polynomials.
The threshold degree of a Boolean function f is the minimum degree of a real polynomial p that represents f in sign: f(x) = sgn p(x). Introduced in the seminal work of Minsky and Papert (1969), this notion is central to some of the strongest algorithmic and complexity-theoretic results for constant-depth circuits. One problem that has remained open for several decades, with applications to computational learning and communication complexity, is to determine the maximum threshold degree of a polynomial-size constant-depth circuit in n variables. The best lower bound prior to our work was Ω(n(d-1)/(2d-1)) for circuits of depth d. We obtain a polynomial improvement for every depth d, with a lower bound of Ω(n3/7) for depth 3 and Ω(√n) for depth d ≥4. The proof contributes a novel approximation-theoretic technique of independent interest, which exploits asymmetry in circuits to prove their hardness for polynomials.
The threshold degree of a Boolean function f is the minimum degree of a real polynomial p that represents f in sign: f(x) ≡ sgn p(x). In a seminal 1969 monograph, Minsky and Papert constructed a polynomial-size constant-depth {∧, ∨)-circuit in n variables with threshold degree Ω(n1/3). This bound underlies some of today's strongest results on constant-depth circuits. It has been an open problem (O'Donnell and Servedio, STOC 2003) to improve Minsky and Papert's bound to nΩ(1)+1/3.
We prove two new results about the inability of low-degree polynomials to uniformly approximate constant-depth circuits, even to slightly-better-than-trivial error. First, we prove a tight Omega~(n^{1/2}) lower bound on the threshold degree of the SURJECTIVITY function on n variables. This matches the best known threshold degree bound for any AC^0 function, previously exhibited by a much more complicated circuit of larger depth (Sherstov, FOCS 2015). Our result also extends to a 2^{Omega~(n^{1/2})} lower bound on the sign-rank of an AC^0 function, improving on the previous best bound of 2^{Omega(n^{2/5})} (Bun and Thaler, ICALP 2016). Second, for any delta>0, we exhibit a function f : {-1,1}^n -> {-1,1} that is computed by a circuit of depth O(1/delta) and is hard to approximate by polynomials in the following sense: f cannot be uniformly approximated to error epsilon=1-2^{-Omega(n^{1-delta})}, even by polynomials of degree n^{1-delta}. Our recent prior work (Bun and Thaler, FOCS 2017) proved a similar lower bound, but which held only for error epsilon=1/3. Our result implies 2^{Omega(n^{1-delta})} lower bounds on the complexity of AC^0 under a variety of basic measures such as discrepancy, margin complexity, and threshold weight. This nearly matches the trivial upper bound of 2^{O(n)} that holds for every function. The previous best lower bound on AC^0 for these measures was 2^{Omega(n^{1/2})} (Sherstov, FOCS 2015). Additional applications in learning theory, communication complexity, and cryptography are described.
The approximate degree of a Boolean function f : {-1, 1}n→ {-1, 1} is the least degree of a real polynomial that approximates f pointwise to error at most 1/3. We introduce a generic method for increasing the approximate degree of a given function, while preserving its computability by constant-depth circuits. Specifically, we show how to transform any Boolean function f with approximate degree d into a function F on O(n · polylog(n)) variables with approximate degree at least D = Ω(n1/3· d2/3). In particular, if d = n1-Ω(1), then D is polynomially larger than d. Moreover, if f is computed by a constant-depth polynomial-size Boolean circuit, then so is F. By recursively applying our transformation, for any constant δ > 0 we exhibit an AC0function of approximate degree Ω(n1-δ). This improves over the best previous lower bound of Ω(n2/3) due to Aaronson and Shi (J. ACM 2004), and nearly matches the trivial upper bound of n that holds for any function. Our lower bounds also apply to (quasipolynomial-size) DNFs of polylogarithmic width. We describe several applications of these results. We give: · For any constant δ > 0, an Ω(n1-δ) lower bound on the quantum communication complexity of a function in AC0. · A Boolean function f with approximate degree at least C(f)2-o(1), where C(f) is the certificate complexity of f. This separation is optimal up to the o(1) term in the exponent. · Improved secret sharing schemes with reconstruction procedures in AC0.
A Nearly Optimal Lower Bound on the Approximate Degree of AC0
The approximate degree of a Boolean function f:{−1,1}n→{−1,1} is the least degree of a real polynomial that approximates f pointwise to error at most 1/3. We introduce a generic method for increasing the approximate degree of a given function, while preserving its computability by constant-depth circuits. Specifically, we show how to transform any Boolean function f with approximate degree d into a function F on O(n⋅polylog(n)) variables with approximate degree at least D=Ω(n1/3⋅d2/3). In particular, if d=n1−Ω(1), then D is polynomially larger than d. Moreover, if f is computed by a polynomial-size Boolean circuit of constant depth, then so is F. By recursively applying our transformation, for any constant δ>0 we exhibit an AC0 function of approximate degree Ω(n1−δ). This improves upon the best previous lower bound of Ω(n2/3) due to Aaronson and Shi [ J. ACM, 51 (2004), pp. 595--605] and nearly matches the trivial upper bound of n that holds for any function. Our lower bounds also apply to (quasipolynomial-size) disjunctive normal forms of polylogarithmic width. We describe several applications of these results and provide the following: (i) for any constant δ>0, an Ω(n1−δ) lower bound on the quantum communication complexity of a function in AC0; (ii) a Boolean function f with approximate degree at least C(f)2−o(1), where C(f) is the certificate complexity of f; this separation is optimal up to the o(1) term in the exponent; (iii) improved secret sharing schemes with reconstruction procedures in AC0.
Abstract. The approximate degree of a Boolean function [Formula: see text] is the minimum degree of a real polynomial [Formula: see text] that approximates [Formula: see text] pointwise: [Formula: see text] for all [Formula: see text]. For every [Formula: see text] we construct conjunctive normal form (CNF) and disjunctive normal form (DNF) formulas of polynomial size with approximate degree [Formula: see text] essentially matching the trivial upper bound of [Formula: see text]. This improves polynomially on previous lower bounds and fully resolves the approximate degree of constant-depth circuits [Formula: see text] a question that has seen extensive research over the past 10 years. Prior to our work, an [Formula: see text] lower bound was known only for [Formula: see text] circuits of depth that grows with [Formula: see text] (M. Bun and J. Thaler, SIAM J. Comput., 49 (2020), pp. FOCS17-59–FOCS17-96). Furthermore, the CNF and DNF formulas that we construct are the simplest possible in that they have constant width. Our proof departs significantly from previous approaches and contributes a novel, number-theoretic method for amplifying approximate degree. Our main result remains valid even for one-sided approximation: for any [Formula: see text], we construct a polynomial-size constant-width CNF formula with one-sided approximate degree [Formula: see text]. We thus obtain the following nearly tight separations: [Formula: see text] versus [Formula: see text] for the one-sided versus two-sided approximate degree of a function and [Formula: see text] versus [Formula: see text] for the one-sided approximate degree of a function [Formula: see text] versus its negation [Formula: see text]. As an application, we essentially settle the communication complexity of [Formula: see text] circuits in the bounded-error quantum model, [Formula: see text]-party number-on-the-forehead randomized model, and [Formula: see text]-party number-on-the-forehead nondeterministic model: we prove that for every [Formula: see text], these models require [Formula: see text], [Formula: see text], and [Formula: see text], respectively, bits of communication even for polynomial-size constant-width CNF formulas. In particular, we show that the multiparty communication class [Formula: see text] can be separated near-optimally from [Formula: see text] and [Formula: see text] by a particularly simple function, a polynomial-size constant-width CNF formula.
We consider the problem of approximating a Boolean function f : f0; 1g n ! f0; 1g by the sign of an integer polynomial p of degree k. For us, a polynomial p(x) predicts the value of f(x) if, whenever p(x) 0, f(x) = 1, and whenever p(x) ! 0, f(x) = 0. A low-degree polynomial p is a good approximator for f if it predicts f at almost all points. Given a positive integer k, and a Boolean function f , we ask, "how good is the best degree k approximation to f?" We introduce a new lower bound technique which applies to any Boolean function. We show that the lower bound technique yields tight bounds in the case f is parity. Minsky and Papert [10] proved that a perceptron can not compute parity; our bounds indicate exactly how well Yale University, Dept. of Computer Science, P.O. Box 208285, New Haven CT 06520-8285. y Email: aspnes-james@cs.yale.edu. z Email: beigel-richard@cs.yale.edu. Supported in part by NSF grants CCR-8808949 and CCR-8958528. x Carnegie-Mellon University, Schoo...