Counting real zeros in the multivariate case

From MaRDI portal





This paper provides a computational base of effective counting the number of a real algebraic set of dimension zero in an algebraic constraint region, generalizing Hermite's theorem to the case of several variables. In the main theorem (theorem 2.1), the number is described by the signature of a bilinear form defined on the structure ring of the algebraic set. A complete proof is given. -- It would be useful to notice that the result is closely related to papers of \textit{D. Eisenbud} and \textit{H. I. Levine} [Ann. Math., II. Ser. 106, 19-44 (1977; Zbl 0398.57020)], \textit{K. Aoki}, \textit{T. Fukuda} and \textit{T. Nishimura} [Topology and computer sciences, Proc. Symp. Atami 1986, 347-363 (1987)], and \textit{J. Montaldi} and \textit{D. van Straten} [Topology 29, No. 4, 501- 510 (1990; Zbl 0755.32032)]. Remarking that the bilinear form is calculated using the Gröbner basis, the authors proceed to study the computational complexity of the algorithm of counting real zeros, based on the main theorem. They also mention the calculation of the Euler characteristic of smooth real hypersurfaces. This paper seems to be readable widely both for pure mathematicians and for specialists of computation. It is therefore a pity that there are several misprints, especially, about the references. For the entire collection see [Zbl 0917.00011].




Cited in
(44)








This page was built for publication: Counting real zeros in the multivariate case

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3135183)