English

Log-concavity and lower bounds for arithmetic circuits

Computational Complexity 2017-01-17 v1 Discrete Mathematics Commutative Algebra

Abstract

One question that we investigate in this paper is, how can we build log-concave polynomials using sparse polynomials as building blocks? More precisely, let f=_i=0da_iXiR+[X]f = \sum\_{i = 0}^d a\_i X^i \in \mathbb{R}^+[X] be a polynomial satisfying the log-concavity condition a_i2\textgreaterτa_i1a_i+1a\_i^2 \textgreater{} \tau a\_{i-1}a\_{i+1} for every i{1,,d1},i \in \{1,\ldots,d-1\}, where τ\textgreater0\tau \textgreater{} 0. Whenever ff can be written under the form f=_i=1k_j=1mf_i,jf = \sum\_{i = 1}^k \prod\_{j = 1}^m f\_{i,j} where the polynomials f_i,jf\_{i,j} have at most tt monomials, it is clear that dktmd \leq k t^m. Assuming that the f_i,jf\_{i,j} have only non-negative coefficients, we improve this degree bound to d=O(km2/3t2m/3log2/3(kt))d = \mathcal O(k m^{2/3} t^{2m/3} {\rm log^{2/3}}(kt)) if τ\textgreater1\tau \textgreater{} 1, and to dkmtd \leq kmt if τ=d2d\tau = d^{2d}. This investigation has a complexity-theoretic motivation: we show that a suitable strengthening of the above results would imply a separation of the algebraic complexity classes VP and VNP. As they currently stand, these results are strong enough to provide a new example of a family of polynomials in VNP which cannot be computed by monotone arithmetic circuits of polynomial size.

Keywords

Cite

@article{arxiv.1503.07705,
  title  = {Log-concavity and lower bounds for arithmetic circuits},
  author = {Ignacio García-Marco and Pascal Koiran and Sébastien Tavenas},
  journal= {arXiv preprint arXiv:1503.07705},
  year   = {2017}
}
R2 v1 2026-06-22T09:02:49.591Z