Robust algorithms for generalized Pham systems (Q626616)

\textit{D. Castro, M. Giusti, J. Heintz, G. Matera} and \textit{L. M. Pardo} [Found. Comput. Math. 3, No. 4, 347--420 (2003; Zbl 1049.68070)] introduced the concept of robustness of an algorithm for an input with parameters. In this paper, the problem for solving generalized Pham systems over the field of the complex numbers is considered. A generalized Pham system is a polynomial system of \(n\) equations in \(n\) variables, for some natural number \(n\), such that the system of the \(n\) leading forms does not have any solutions except zero. The Bezout number of a generalized Pham system is the number of solutions counted with multiplicity. To solve a system means to give a univariate polynomial whose solutions correspond to the solution of the system, and a polynomial map from the set of roots of the polynomial to the set of solutions of the system. The paper gives and upper and a lower bound for the complexity of this problem, in terms of the Bezout number. The model of computation is based on the concept of straight-line programs.