Waring and cactus ranks and strong Lefschetz property for annihilators of symmetric forms (Q2114119)

From MaRDI portal





scientific article
Language Label Description Also known as
default for all languages
No label defined
    English
    Waring and cactus ranks and strong Lefschetz property for annihilators of symmetric forms
    scientific article

      Statements

      Waring and cactus ranks and strong Lefschetz property for annihilators of symmetric forms (English)
      0 references
      0 references
      0 references
      0 references
      0 references
      15 March 2022
      0 references
      Let \(k\) be a field of characteristic 0, let \(S=k[x_1,\ldots,x_n]\) be a polynomial ring and let \(E=k[X_1,\ldots,X_n]\) be its inverse system. It means that \(E\) is a \(S\)-module where \(x_i\) acts as \(\partial /\partial X_i\). The first result of the paper is a proof that the Gorenstein algebra \(S/\mathrm{ann}(h_{n,d})\) is compressed and satisfies the Strong Lefschetz Property (SLP), where \[h_{n,d}(X_1,\ldots,X_n)=\sum_{\underline{i}\in\mathbb{N}^n,|i|=d} X_1^{i_1}\cdot\ldots\cdot X_n^{i_n}\] is the complete symmetric polynomial. Recal that a graded artinian \(k\)-algebra \(A\) \textit{satisfies the SLP} if a multiplication by a power of general linear form \[\times l^j: [A]_i\rightarrow [A]_{i+j}\] for all \(i\geq 0\), \(j\geq 0\) has maximal rank. The authors give also a useful decomposition of \(\mathrm{ann}(h_{n,d})\), separating into the cases where \(d\) is even or odd. In the next sections the paper concerns the Waring rank. If \(f\) is a homogeneous polynomial of degree \(m\), then its \textit{Waring rank} is the smallest \(k\) such that \(f\) can be written in the form \[L_1^m+\ldots +L_k^m\] for linear forms \(L_1, \ldots, L_k\), and \textit{power sum decomposition} is any expression of \(f\) in this form. The authors give a decomposition of the complete linear form, separating into cases where \(d\) is even or odd. In both cases the power sum decomposition has \(\binom{n+d}{d}\) terms, giving an upper bound for the Waring rank. This bound is a polynomial of degree \(d\) in \(n\). They show that this bound for symmetric cubic polynomial is sharp. Next the authors study the artinian Gorenstein algebras defined by the annihilator of symmetric cubic forms in \(n\geq 3\) variables. They determine the possible Hilbert function of the annihilator, the Waring rank, tha cactus rank, and the resolution of all such Gorenstein algebras. Recall thet the \textit{cactus rank} of a form \(F\) is the smallest degree of a zero-dimensional subscheme whose saturated ideal is contained in \(\mathrm{ann}(F)\). As a consequence they show that for symmetric cubic forms, the Waring rank and the cactus rank are equal expect for point on a line tangent to the cubic at the cusp \(P\) other then the cusp. They also show that for \textit{any} symmetric cubic form \(F\), the corresponding algebra \(S/\mathrm{ann}(F)\) satisfies the SLP. In the last section the authors give upper bounds for generic Waring rank (it means the Waring rank of a generic symmetric form) for quartics and quintics.
      0 references
      Waring rank
      0 references
      cactus rank
      0 references
      symmetric forms
      0 references
      strong Lefschetz property
      0 references
      Macaulay duality
      0 references
      minimal free resolution
      0 references
      power sum decomposition
      0 references
      Gorenstein algebra
      0 references

      Identifiers