Representations of positive polynomials on noncompact semialgebraic sets via KKT ideals

This paper studies the representation of a positive polynomial f(x) on a noncompact semialgebraic set S={x∈Rⁿ:g₁(x)≥0,…,g_s(x)≥0} modulo its KKT (Karush-Kuhn-Tucker) ideal. Under the assumption that the minimum value of f(x) on S is attained at some KKT point, we show that f(x) can be represented as sum of squares (SOS) of polynomials modulo the KKT ideal if f(x)>0 on S; furthermore, when the KKT ideal is radical, we argue that f(x) can be represented as a sum of squares (SOS) of polynomials modulo the KKT ideal if f(x)≥0 on S. This is a generalization of results in [J. Nie, J. Demmel, B. Sturmfels, Minimizing polynomials via sum of squares over the gradient ideal, Mathematical Programming (in press)], which discusses the SOS representations of nonnegative polynomials over gradient ideals.     

Strict positivstellensätze for matrix polynomials with scalar constraints

We extend Krivine’s strict positivstellensätz for usual (real multivariate) polynomials to symmetric matrix polynomials with scalar constraints. The proof is an elementary computation with Schur complements. Analogous extensions of Schmüdgen’s and Putinar’s strict positivstellensätz were recently proved by Hol and Scherer using methods from optimization theory.     

A note on non-reduced Picard schemes

The Picard scheme of a smooth curve and a smooth complex variety is reduced. In this note we discuss which classes of surfaces in terms of the Enriques-Kodaira classification can have non-reduced Picard schemes and whether there are restrictions on the characteristic of the ground field. It turns out that non-reduced Picard schemes are uncommon in Kodaira dimension κ≤0, that this phenomenon can be bounded for κ=2 (general type) and that it is as bad as can be for κ=1.     

On semidefinite representations of non-closed sets

Spectrahedra are sets defined by linear matrix inequalities. Projections of spectrahedra are called semidefinitely representable sets. Both kinds of sets are of practical use in polynomial optimization, since they occur as feasible sets in semidefinite programming. There are several recent results on the question which sets are semidefinitely representable. So far, all results focus on the case of closed sets.In this work we develop a new method to prove semidefinite representability of sets which are not closed. For example, the interior of a semidefinitely representable set is shown to be semidefinitely representable. More general, one can remove faces of a semidefinitely representable set and preserve semidefinite representability, as long as the faces are parametrized in a suitable way.     

Semidefinite programming and sums of hermitian squares of noncommutative polynomials

An algorithm for finding sums of hermitian squares decompositions for polynomials in noncommuting variables is presented. The algorithm is based on the “Newton chip method”, a noncommutative analog of the classical Newton polytope method, and semidefinite programming.     

Varieties fibred over abelian varieties with fibres of log general type

Let (X,B)$(X,B)$ be a complex projective klt pair, and let f:X→Z$f:X\to Z$ be a surjective morphism onto a normal projective variety with maximal albanese dimension such that _KX+B$K_X+B$ is relatively big over Z. We show that such pairs have good log minimal models.     

Addendum to “Connesʼ embedding conjecture and sums of hermitian squares” [Adv. Math. 217 (4) (2008) 1816–1837]

We show that Connes? embedding conjecture (CEC) is equivalent to a real version of the same (RCEC). Moreover, we show that RCEC is equivalent to a real, purely algebraic statement concerning trace positive polynomials. This purely algebraic reformulation of CEC had previously been given in both a real and a complex version in a paper of the last two authors. The second author discovered a gap in this earlier proof of the equivalence of CEC to the real algebraic reformulation (the proof of the complex algebraic reformulation being correct). In this note, we show that this gap can be filled with help of the theory of real von Neumann algebras.     

The n-generator property in rings of integer-valued polynomials determined by finite sets

Let D be an integral domain and E a non-empty finite subset of D. For n ≧ 2, we show that D has the n-generator property if and only if Int(E, D) has the n-generator property if and only if Int(E, D) has the strong (n + 1)-generator property. Thus, iterating the Int(E, D) construction cannot produce Prüfer domains whose finitely generated ideals require an ever larger number of generators. We also show that, for n ≧ 2, a non-zero polynomial f ∈Int(E, D) is a strong n-generator in Int(E, D) if and only if f (a) is a strong n-generator in D for all a ∈E. Received: 15 July 2004     

A note on threefolds with a hyperplane section of general type

This paper deals with polarized pairs , where is a nonsingular projective threefold and is a very ample line bundle on it, such that for one smooth member  | |, one has (Â)=2. A large class of pairs whose adjoint line bundle is nef and big was indirectedly studied by Beltrametti and co-workers. We add some more information, both in this general case and also when the adjoint line bundle fails to be nef and big.     

A nondensity property of preperiodic points on Chebyshev dynamical systems

Let k be a number field with algebraic closure , and let S be a finite set of primes of k, containing all the infinite ones. Consider a Chebyshev dynamical system on P². Fix the effective divisor D of P² that is equal to a line nondegenerate on²[−2,2]. Then we will prove that the set of preperiodic points on which are S-integral relative to D is not Zariski dense in P².     

Approximation by complex genuine Durrmeyer type polynomials in compact disks

In this paper, the order of simultaneous approximation and Voronovskaja kind results with quantitative estimate for the complex genuine Durrmeyer polynomials attached to analytic functions on compact disks are obtained. In this way, we put in evidence the overconvergence phenomenon for the genuine Durrmeyer polynomials, namely the extensions of the approximation properties (with quantitative estimates) from real intervals to compact disks in the complex plane.     

Remarks on families of singular curves with hyperelliptic normalizations

We give restrictions on the existence of families of curves on smooth projective surfaces S of nonnegative Kodaira dimension all having constant geometric genus p_g ? 2 and hyperelliptic normalizations. In particular, we prove a Reider-like result that relies on deformation theory and bending-and-breaking of rational curves in Sym²(S). We also give examples of families of such curves.     

Ring structures on groups of arithmetic functions

Using the theory of Witt vectors, we define ring structures on several well-known groups of arithmetic functions, which in another guise are formal Dirichlet series. The set of multiplicative arithmetic functions over a commutative ring R is shown to have a unique functorial ring structure for which the operation of addition is Dirichlet convolution and the operation of multiplication restricted to the completely multiplicative functions coincides with point-wise multiplication. The group of additive arithmetic functions over R also has a functorial ring structure. In analogy with the ghost homomorphism of Witt vectors, there is a functorial ring homomorphism from the ring of multiplicative functions to the ring of additive functions that is an isomorphism if R is a Q-algebra. The group of rational arithmetic functions, that is, the group generated by the completely multiplicative functions, forms a subring of the ring of multiplicative functions. The latter ring has the structure of a Bin(R)-algebra, where Bin(R) is the universal binomial ring equipped with a ring homomorphism to R. We use this algebra structure to study the order of a rational arithmetic function, as well the powersfα for α∈Bin(R) of a multiplicative arithmetic function f. For example, we prove new results about the powers of a given multiplicative arithmetic function that are rational. Finally, we apply our theory to the study of the zeta function of a scheme of finite type over Z.     

A remark on the moduli field of a curve

We give a purely algebro-geometric proof of the fact that every nonsingular projective curve can be defined over a finite extension of its moduli field. This extends a result byWolfart [7] to curves over fields of arbitrary characteristic. Received: 30 November 2001     

Rational classes and divisors on curves of genus 2

We describe a method of looking for rational divisor classes on a curve of genus 2. We have an algorithm to decide if a given class of divisors of degree 3 contains a rational divisor. It is known that the shape of the kernel of Cassel’s morphism (X − T) is related to the existence of rational classes of degree 1. Our key tool is the dual Kummer surface.V. G. L. Neumann supported by CNPq, Brazil     

On Brown’s constant associated with irreducible polynomials over henselian valued fields

Let v be a henselian valuation of arbitrary rank of a field K and be the prolongation of v to the algebraic closure of K with value group . In 2008, Ron Brown gave a class P of monic irreducible polynomials over K such that to each g(x) belonging to P, there corresponds a smallest constant λ_g belonging to (referred to as Brown’s constant) with the property that whenever is more than λ_g with K(β) a tamely ramified extension of (K,v), then K(β) contains a root of g(x). In this paper, we determine explicitly this constant besides giving an important property of λ_g without assuming that K(β)/K is tamely ramified.     

Epimorphic extensions and Prüfer extensions¶ of partially ordered rings

The paper contributes to the investigation of epimorphisms in the category of reduced partially ordered rings (porings). Two main questions are considered: 1) Does the set of isomorphism classes of a given poring have a largest element (an epimorphic hull)? 2) Given an epimorphic extension, or even a Prüfer extension, f:A→B of porings: how closely are A and B related to each other? Received: 17 June 1999     

On the notion of canonical dimension for algebraic groups

We define and study a numerical invariant of an algebraic group action which we call the canonical dimension. We then apply the resulting theory to the problem of computing the minimal number of parameters required to define a generic hypersurface of degree d in P^n-1.