Searcharxiv⌕ Search

arXiv subjects

Éric Schost

Publications and source records attributed to Éric Schost.

At least 19 recordsLinked to original sources

Fast Deterministic Normal Bases and Circulant Polynomial Determinants

Let $\mathsf{E}=\mathbb{F}_q[x]/(Γ)$ describe an algebraic extension of a finite field $\mathbb{F}_q$, where $q$ is a prime power and $Γ\in\mathbb{F}_q[x]$ is monic and irreducible of degree $n$. We give a deterministic algorithm that finds $β\in \mathsf{E}$ whose conjugates $β,β^q,\ldots,β^{q^{n-1}}$ form an $\mathbb{F}_q$-basis of $\mathsf{E}/\mathbb{F}_q$, a normal basis, using $O_ε((n^2\log q)^{1+ε})+{O\tilde{}}(n\log^2 q)$ bit operations for any $ε>0$. For $n>1$, let $θ=x\bmodΓ$, so $\mathsf{E}=\mathbb{F}_q[θ]$. A variant of a construction of Artin shows that $β_t=(θ-t)^{-1}$ is normal for all but at most $n(n-1)$ parameters $t\in\mathbb{F}_q$. We present an algorithm to construct an $n\times n$ circulant matrix over $\mathbb{F}_q[\mathcal T]$, for an indeterminate $\mathcal T$, whose determinant at $\mathcal T=t$ is non-zero precisely when $β_t$ is normal, and show this algorithm requires ${O\tilde{}}(n^2+n\log q)$ operations in $\mathbb{F}_q$. Then, as a primary subroutine, using triangular-set power projection and modular composition, we show how to compute the determinant of any $n\times n$ circulant over $\mathbb{F}_q[\mathcal T]$, given by its first row of polynomials of degree at most $m\geq1$, using $O_ε((nm\log q)^{1+ε})$ bit operations. For $q\leq n(n-1)$, we show how to embed our problem into a sufficiently large field extension, construct a normal basis there, and descend to the ground field, within the same stated cost.

cs.SC↗

Faster modular composition using two relation matrices

Modular composition is the problem of computing the composition of two univariate polynomials modulo a third one. For a long time, the fastest algebraic algorithm for this problem was that of Brent and Kung (1978). Recently, we improved Brent and Kung's algorithm by computing and using a polynomial matrix that encodes a certain basis of algebraic relations between the polynomials. This is further improved here by making use of two polynomial matrices of smaller dimension. Under genericity assumptions on the input, this results in an algorithm using $\tilde{O}(n^{(ω+3)/4})$ arithmetic operations in the base field, where $ω$ is the exponent of matrix multiplication. With naive matrix multiplication, this is $\tilde{O}(n^{3/2})$, while with the best currently known exponent $ω$ this is $O(n^{1.343})$, improving upon the previously most efficient algorithms.

cs.SC↗

Refined bit complexity for the computation of at least one point per connected component of a smooth complete intersection real algebraic set

We refine the bit complexity analysis of an algorithm for the computation of at least one point per connected component of a smooth real algebraic set, yielding exponential speedup (with respect to the number of variables) compared to prior works. The algorithm which is analyzed is based on the critical point method, reducing the problem to computations of critical points associated to the restriction of generic projections on lines to the studied variety. Our refinement, and the subsequent improved complexity statement, comes from a better utilization of the multi-affine structure of polynomial systems encoding these sets of critical points. The bit-size estimates on the size of the output produced by this algorithm are also improved by this refinement.

cs.SC↗

Primes of bad reduction for systems of polynomial equations

Consider polynomials $F_1,\dots,F_s$ in $\K[X_1,\dots,X_n]$ over a field $\K$, their zero-set $V(F_1,\dots,F_n)$ in $\Kbar^n$ and its decomposition into equidimensional components $V_0,\dots,V_n$ (with $V_i$ either empty or of dimension $i$ for all $i$). To each $V_i$, we can associate its Chow forms, which are polynomials in new variables $(U_{k,j})_{0\le k\le i, 0 \le j \le n}$, uniquely defined up to a scalar factor. These Chow forms completely characterize $V_i$: we can recover equations for $V_i$ from them, and their degree is $(i+1)$ times the degree of $V_i$. We discuss the situation when the $F_i$'s have integer coefficients, and study the question of when the Chow forms of the $V_i$'s defined as above can be reduced modulo $p$ to give Chow forms of the equidimensional components of $V(F_1 \bmod p,\dots,F_s \bmod p)$. We show that this is the case as soon as $p$ does not divide a certain nonzero integer $Δ$ of height $O(n^{14} s h d^{3n+4})$, with $d$ and $h$ bounds on respectively the degrees and heights of the $F_i$'s.

math.AC↗

Hilbert's Nullstellensatz is in the Counting Hierarchy

We show that Hilbert's Nullstellensatz, the problem of deciding if a system of multivariate polynomial equations has a solution in the algebraic closure of the underlying field, lies in the counting hierarchy. More generally, we show that the number of solutions to a system of equations can be computed in polynomial time with oracle access to the counting hierarchy. Our results hold in particular for polynomials with coefficients in either the rational numbers or a finite field. Previously, the best-known bounds on the complexities of these problems were PSPACE and FPSPACE, respectively. Our main technical contribution is the construction of a uniform family of constant-depth arithmetic circuits that compute the multivariate resultant.

cs.CC↗

Computing roadmaps in unbounded smooth real algebraic sets II: algorithm and complexity

A roadmap for an algebraic set $V$ defined by polynomials with coefficients in the field $\mathbb{Q}$ of rational numbers is an algebraic curve contained in $V$ whose intersection with all connected components of $V\cap\mathbb{R}^{n}$ is connected. These objects, introduced by Canny, can be used to answer connectivity queries over $V\cap \mathbb{R}^{n}$ provided that they are required to contain the finite set of query points $\mathcal{P}\subset V$; in this case, we say that the roadmap is associated to $(V, \mathcal{P})$. In this paper, we make effective a connectivity result we previously proved, to design a Monte Carlo algorithm which, on input (i) a finite sequence of polynomials defining $V$ (and satisfying some regularity assumptions) and (ii) an algebraic representation of finitely many query points $\mathcal{P}$ in $V$, computes a roadmap for $(V, \mathcal{P})$. This algorithm generalizes the nearly optimal one introduced by the last two authors by dropping a boundedness assumption on the real trace of $V$. The output size and running times of our algorithm are both polynomial in $(nD)^{n\log d}$, where $D$ is the maximal degree of the input equations and $d$ is the dimension of $V$. As far as we know, the best previously known algorithm dealing with such sets has an output size and running time respectively polynomial in $(n^{\log{n}}D)^{n\log n}$ and $(n^{\log{n}}D)^{n\log^2 n}$.

cs.SC↗

A SageMath Package for Analytic Combinatorics in Several Variables: Beyond the Smooth Case

The field of analytic combinatorics in several variables (ACSV) develops techniques to compute the asymptotic behaviour of multivariate sequences from analytic properties of their generating functions. When the generating function under consideration is rational, its set of singularities forms an algebraic variety -- called the singular variety -- and asymptotic behaviour depends heavily on the geometry of the singular variety. By combining a recent algorithm for the Whitney stratification of algebraic varieties with methods from ACSV, we present the first software that rigorously computes asymptotics of sequences whose generating functions have non-smooth singular varieties (under other assumptions on local geometry). Our work is built on the existing sage_acsv package for the SageMath computer algebra system, which previously gave asymptotics under a smoothness assumption. We also report on other improvements to the package, such as an efficient technique for determining higher order asymptotic expansions using Newton iteration, the ability to use more efficient backends for algebraic computations, and a method to compute so-called critical points for any multivariate rational function through Whitney stratification.

math.CO↗

Faster Modular Composition

A new Las Vegas algorithm is presented for the composition of two polynomials modulo a third one, over an arbitrary field. When the degrees of these polynomials are bounded by $n$, the algorithm uses $O(n^{1.43})$ field operations, breaking through the $3/2$ barrier in the exponent for the first time. The previous fastest algebraic algorithms, due to Brent and Kung in 1978, require $O(n^{1.63})$ field operations in general, and ${n^{3/2+o(1)}}$ field operations in the special case of power series over a field of large enough characteristic. If cubic-time matrix multiplication is used, the new algorithm runs in ${n^{5/3+o(1)}}$ operations, while previous ones run in $O(n^2)$ operations. Our approach relies on the computation of a matrix of algebraic relations that is typically of small size. Randomization is used to reduce arbitrary input to this favorable situation.

cs.SC↗

Computing roadmaps in unbounded smooth real algebraic sets I: connectivity results

Answering connectivity queries in real algebraic sets is a fundamental problem in effective real algebraic geometry that finds many applications in e.g. robotics where motion planning issues are topical. This computational problem is tackled through the computation of so-called roadmaps which are real algebraic subsets of the set V under study, of dimension at most one, and which have a connected intersection with all semi-algebraically connected components of V. Algorithms for computing roadmaps rely on statements establishing connectivity properties of some well-chosen subsets of V , assuming that V is bounded. In this paper, we extend such connectivity statements by dropping the boundedness assumption on V. This exploits properties of so-called generalized polar varieties, which are critical loci of V for some well-chosen polynomial maps.

cs.SC↗

Faster real root decision algorithm for symmetric polynomials

In this paper, we consider the problem of deciding the existence of real solutions to a system of polynomial equations having real coefficients, and which are invariant under the action of the symmetric group. We construct and analyze a Monte Carlo probabilistic algorithm which solves this problem, under some regularity assumptions on the input, by taking advantage of the symmetry invariance property. The complexity of our algorithm is polynomial in $d^s, {{n+d} \choose d}$, and ${{n} \choose {s+1}}$, where $n$ is the number of variables and $d$ is the maximal degree of $s$ input polynomials defining the real algebraic set under study. In particular, this complexity is polynomial in $n$ when $d$ and $s$ are fixed and is equal to $n^{O(1)}2^n$ when $d=n$.

cs.SC↗

Computing the Characteristic Polynomial of Endomorphisms of a finite Drinfeld Module using Crystalline Cohomology

We present a new algorithm for computing the characteristic polynomial of an arbitrary endomorphism of a finite Drinfeld module using its associated crystalline cohomology. Our approach takes inspiration from Kedlaya's p-adic algorithm for computing the characteristic polynomial of the Frobenius endomorphism on a hyperelliptic curve using Monsky-Washnitzer cohomology. The method is specialized using a baby-step giant-step algorithm for the particular case of the Frobenius endomorphism, and in this case we include a complexity analysis that demonstrates asymptotic gains over previously existing approaches

cs.SC↗

Newton iteration for lexicographic Gröbner bases in two variables

We present an $m$-adic Newton iteration with quadratic convergence for lexicographic Gröbner basis of zero dimensional ideals in two variables. We rely on a structural result about the syzygies in such a basis due to Conca and Valla, that allowed them to explicitly describe these Gröbner bases by affine parameters; our Newton iteration works directly with these parameters.

math.AC↗

Algorithms for Linearly Recurrent Sequences of Truncated Polynomials

Linear recurrent sequences are those whose elements are defined as linear combinations of preceding elements, and finding recurrence relations is a fundamental problem in computer algebra. In this paper, we focus on sequences whose elements are vectors over the ring $\mathbb{A} = \mathbb{K}[x]/(x^d)$ of truncated polynomials. Finding the ideal of their recurrence relations has applications such as the computation of minimal polynomials and determinants of sparse matrices over $\mathbb{A}$. We present three methods for finding this ideal: a Berlekamp-Massey-like approach due to Kurakin, one which computes the kernel of some block-Hankel matrix over $\mathbb{A}$ via a minimal approximant basis, and one based on bivariate Padé approximation. We propose complexity improvements for the first two methods, respectively by avoiding the computation of redundant relations and by exploiting the Hankel structure to compress the approximation problem. Then we confirm these improvements empirically through a C++ implementation, and we discuss the above-mentioned applications.

cs.SC↗

Subquadratic-Time Algorithms for Normal Bases

For any finite Galois field extension $\mathsf{K}/\mathsf{F}$, with Galois group $G = \mathrm{Gal}(\mathsf{K}/\mathsf{F})$, there exists an element $α\in \mathsf{K}$ whose orbit $G\cdotα$ forms an $\mathsf{F}$-basis of $\mathsf{K}$. Such a $α$ is called a normal element and $G\cdotα$ is a normal basis. We introduce a probabilistic algorithm for testing whether a given $α\in \mathsf{K}$ is normal, when $G$ is either a finite abelian or a metacyclic group. The algorithm is based on the fact that deciding whether $α$ is normal can be reduced to deciding whether $\sum_{g \in G} g(α)g \in \mathsf{K}[G]$ is invertible; it requires a slightly subquadratic number of operations. Once we know that $α$ is normal, we show how to perform conversions between the power basis of $\mathsf{K}/\mathsf{F}$ and the normal basis with the same asymptotic cost.

cs.SC↗

Homotopy techniques for solving sparse column support determinantal polynomial systems

Let $\mathbf{K}$ be a field of characteristic zero with $\overline{\mathbf{K}}$ its algebraic closure. Given a sequence of polynomials $\mathbf{g} = (g_1, \ldots, g_s) \in \mathbf{K}[x_1, \ldots , x_n]^s$ and a polynomial matrix $\mathbf{F} = [f_{i,j}] \in \mathbf{K}[x_1, \ldots, x_n]^{p \times q}$, with $p \leq q$, we are interested in determining the isolated points of $V_p(\mathbf{F},\mathbf{g})$, the algebraic set of points in $\overline{\mathbf{K}}$ at which all polynomials in $\mathbf{g}$ and all $p$-minors of $\mathbf{F}$ vanish, under the assumption $n = q - p + s + 1$. Such polynomial systems arise in a variety of applications including for example polynomial optimization and computational geometry. We design a randomized sparse homotopy algorithm for computing the isolated points in $V_p(\mathbf{F},\mathbf{g})$ which takes advantage of the determinantal structure of the system defining $V_p(\mathbf{F}, \mathbf{g})$. Its complexity is polynomial in the maximum number of isolated solutions to such systems sharing the same sparsity pattern and in some combinatorial quantities attached to the structure of such systems. It is the first algorithm which takes advantage both on the determinantal structure and sparsity of input polynomials. We also derive complexity bounds for the particular but important case where $\mathbf{g}$ and the columns of $\mathbf{F}$ satisfy weighted degree constraints. Such systems arise naturally in the computation of critical points of maps restricted to algebraic sets when both are invariant by the action of the symmetric group.

cs.SC↗

Computing critical points for invariant algebraic systems

Let $\mathbf{K}$ be a field and $ϕ$, $\mathbf{f} = (f_1, \ldots, f_s)$ in $\mathbf{K}[x_1, \dots, x_n]$ be multivariate polynomials (with $s < n$) invariant under the action of $\mathcal{S}_n$, the group of permutations of $\{1, \dots, n\}$. We consider the problem of computing the points at which $\mathbf{f}$ vanish and the Jacobian matrix associated to $\mathbf{f}, ϕ$ is rank deficient provided that this set is finite. We exploit the invariance properties of the input to split the solution space according to the orbits of $\mathcal{S}_n$. This allows us to design an algorithm which gives a triangular description of the solution space and which runs in time polynomial in $d^s$, ${{n+d}\choose{d}}$ and $\binom{n}{s+1}$ where $d$ is the maximum degree of the input polynomials. When $d,s$ are fixed, this is polynomial in $n$ while when $s$ is fixed and $d \simeq n$ this yields an exponential speed-up with respect to the usual polynomial system solving algorithms.

cs.SC↗

Computing syzygies in finite dimension using fast linear algebra

We consider the computation of syzygies of multivariate polynomials in a finite-dimensional setting: for a $\mathbb{K}[X_1,\dots,X_r]$-module $\mathcal{M}$ of finite dimension $D$ as a $\mathbb{K}$-vector space, and given elements $f_1,\dots,f_m$ in $\mathcal{M}$, the problem is to compute syzygies between the $f_i$'s, that is, polynomials $(p_1,\dots,p_m)$ in $\mathbb{K}[X_1,\dots,X_r]^m$ such that $p_1 f_1 + \dots + p_m f_m = 0$ in $\mathcal{M}$. Assuming that the multiplication matrices of the $r$ variables with respect to some basis of $\mathcal{M}$ are known, we give an algorithm which computes the reduced Gröbner basis of the module of these syzygies, for any monomial order, using $O(m D^{ω-1} + r D^ω\log(D))$ operations in the base field $\mathbb{K}$, where $ω$ is the exponent of matrix multiplication. Furthermore, assuming that $\mathcal{M}$ is itself given as $\mathcal{M} = \mathbb{K}[X_1,\dots,X_r]^n/\mathcal{N}$, under some assumptions on $\mathcal{N}$ we show that these multiplication matrices can be computed from a Gröbner basis of $\mathcal{N}$ within the same complexity bound. In particular, taking $n=1$, $m=1$ and $f_1=1$ in $\mathcal{M}$, this yields a change of monomial order algorithm along the lines of the FGLM algorithm with a complexity bound which is sub-cubic in $D$.

cs.SC↗

Computing the Characteristic Polynomial of a Finite Rank Two Drinfeld Module

Motivated by finding analogues of elliptic curve point counting techniques, we introduce one deterministic and two new Monte Carlo randomized algorithms to compute the characteristic polynomial of a finite rank-two Drinfeld module. We compare their asymptotic complexity to that of previous algorithms given by Gekeler, Narayanan and Garai-Papikian and discuss their practical behavior. In particular, we find that all three approaches represent either an improvement in complexity or an expansion of the parameter space over which the algorithm may be applied. Some experimental results are also presented.

cs.SC↗