Products of polynomials in many variables (Q805783)

From MaRDI portal
Revision as of 11:05, 30 January 2024 by Import240129110113 (talk | contribs) (Added link to MaRDI item.)
scientific article
Language Label Description Also known as
English
Products of polynomials in many variables
scientific article

    Statements

    Products of polynomials in many variables (English)
    0 references
    0 references
    0 references
    0 references
    0 references
    1990
    0 references
    Let P and Q denote polynomials in N variables, with complex coefficients. The authors are interested in estimates of the form \(\| PQ\| \geq \lambda \| P\| \cdot \| Q\|\), where \(\| \cdot \|\) is some norm on the space of polynomials. There is particular interest in the case that \(\lambda\) does not depend on the number of variables. The norms considered are \(| P|_ p\), the \(\ell_ p\) norm of the coefficients of P, and \(\| P\|_ p\) the \(L_ p\) norm of P on the polycircle (N-torus). For these norms, they show (Theorems 2.1 and 2.2) that \(\| PQ\| /(\| P\| \cdot \| Q\|)\) is bounded above and below by (finite and positive) constants which depend on the (total) degrees of P and Q and on the norm in question. Let \(P|^ m\) denote the projection of P onto polynomials of (total) degree at most m. Then P is said to have concentration d at degree m in norm \(\| \cdot \|\) if \(\| P|^ m\| \geq d\| P\|\). Theorems 2.3 and 2.5 show that if \(0<d\), \(d'\leq 1\) and if P, Q have concentrations d, \(d'\) at degrees m, n in norm \(\| \cdot \|\), then there is a constant \(\lambda >0\) depending on d, \(d'\), m, n and the norm \(\| \cdot \|\) (but not on the degrees of P and Q, nor on the number of variables), such that \(\| PQ\| \geq \lambda \| P\| \cdot \| Q\|.\) Generalizing a result of the first and third author for polynomials in one variable [J. Number Theory 21, 390-412 (1985; Zbl 0578.30004)], they show that if P has a large coefficient and Q has a large coefficient at low degrees, then the product PQ has a large coefficient, with estimates independent of the number of variables (Theorem 3.1). The authors employ a wide variety of interesting methods to prove their results.
    0 references
    0 references
    norm on the space of polynomials
    0 references
    degree
    0 references
    concentration
    0 references

    Identifiers