Manfred Minimair
Computing Resultants of Partially Composed Polynomials
Computer Algebra in Scientific Computing.
Proceedings of the CASC 2004 (St. Petersburg, Russia)
(Editors: V. G. Ganzha, E. W. Mayr and E. V. Vorozhtsov)
TUM Muenchen, 359-366, 2004

Abstract

This paper is part of the author's work on determining the irreducible factors of sparse (toric) resultants of composed polynomials. The motivation behind this work is to use the factors for efficient elimination of variables, by sparse resultant computation, from composed polynomials. Previous works considered the sparse (toric) resultant of polynomials having arbitrary (mixed) supports composed with (i.e. evaluated at) polynomials having the same (unmixed) supports and of polynomials having the same (unmixed) supports composed with polynomials having arbitrary (mixed) supports, resp. Here, we consider the sparse resultant of linear polynomials having arbitrary (mixed) supports composed with polynomials having arbitrary (mixed) supports, also called ``linearly combined polynomials'', (under a natural assumption on their exponents).

The main contribution of this paper is to determine the irreducible factors, together with their exponents, of the sparse resultant of these linearly combined polynomials. This result essentially generalizes a result by Gelfand, Kapranov and Zelevinsky factoring the sparse resultant of unmixed dense linear polynomials composed with polynomials with unmixed supports. It is expected that this result can be applied to eliminate variables from linearly combined polynomials with improved efficiency.

Preprint