Effective Nullstellensatz for arbitrary ideals

Polynomials in number theory (11C08) Polynomials over commutative rings (13B25) Relevant commutative algebra (14A05) Intersection theory, characteristic classes, intersection multiplicities in algebraic geometry (14C17) Formal methods and deformations in algebraic geometry (14D15) Effectivity, complexity and computational aspects of algebraic geometry (14Q20)

Abstract: Let

f_{i}

be polynomials in

n

variables without a common zero. Hilbert's Nullstellensatz says that there are polynomials

g_{i}

such that

s u m g_{i} f_{i} = 1

. The effective versions of this result bound the degrees of the

g_{i}

in terms of the degrees of the

f_{j}

. The aim of this paper is to generalize this to the case when the

f_{i}

are replaced by arbitrary ideals. Applications to the B'ezout theorem, to L ojasiewicz--type inequialities and to deformation theory are also discussed.

Review text

Hilbert's Nullstellensatz states that, given \(f_1,\dots,f_s\) in a ring of polynomials over a field with no common zero in an algebraic closure of this field, there exist \(g_1,\dots,g_s\) in the same ring of polynomials such that \(1=\sum^s_{i=1} g_if_i\). The aim of effective versions of this theorem is to bound the degrees of the polynomials \(g_1,\dots,g_s\) in terms of invariants related to the polynomials \(f_1,\dots,f_s\) (for example, in terms of their degrees and the number of variables involved). In case the polynomials do share common zeroes, there are other related questions one can ask about the intersection of the hypersurfaces they define, such as an effective algebraic Bézout theorem and an effective Łojasiewicz inequality (when the base field is \(\mathbb{C})\). In this paper, the author generalizes these questions to the situation ideals (and not just polynomials). The three main results, adapted to the case the ideals are unmixed, are the following ones: 1. Algebraic Bézout theorem: Let \(K\) be any field and \(I_1,\dots,I_m\) unmixed ideals in \(K[x_1,\dots,x_n]\). Then, there are prime ideals \(P_j\supset (I_1, \dots, I_m)\) and natural numbers \(a_j\) such that \(\prod_jP_j^{a_j} \subset(I_1, \dots,I_m)\) and that \(\sum_ja_j\leq n\cdot\prod_i\deg I_i\). 2. Effective Nullstellensatz: Let \(K\) be any field and \(I_1,\dots,I_m\) unmixed ideals in \(K[x_1, \dots,x_n]\). Then \(I_1,\dots,I_m\) have no common zero in an algebraic closure of \(K\) if and only if there are polynomials \(f_j\in I_j\) such that \(\sum_n f_j=1\) satisfying \(\deg f_i\leq (n+1)\cdot\prod_i\deg I_i\). 3. Effective Łojasiewicz inequality: Let \(I_1,\dots,I_m\) be unmixed ideals in \(\mathbb{C}[x_1, \dots,x_n]\) and let \(X_1,\dots, X_m\subset \mathbb{C}^n\) be the corresponding subschemes. Let \(f_{ij}\) be generators of \(I_i\). Then, for every bounded set \(B\subset\mathbb{C}^n\) there is a real constant \(C>0\) such that, for every \(x\in B\), \(\text{dist} (X_1\cap\cdots \cap X_m,x)^{\Pi_i\deg I_i}< C. \max_{i,j}|f_{ij}(x) |\). The proofs of these theorems (in fact, they are stated in a more general case without the unmixedness hypothesis, and the bounds have to do with the arithmetic degree of the ideals involved) use facts from cycle intersection theory in \(\mathbb{A}^n\), integral closure of ideals and the ideal of Chow equations. In the last section, the author shows an application of his results to deformation theory.

Recommendations

Cites work

Cited in

(34)

This page was built for publication: Effective Nullstellensatz for arbitrary ideals

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1963839)