Numerical Primary Decomposition

Anton Leykin (Institute for Mathematics and Its Applications)

Abstract:

We present a concept that is a numerical counterpart of a primary decomposition of a polynomial ideal in the usual sense. Our algorithm for the numerical primary decomposition is based on a blowing up procedure that stems from the higher-order deflation method for regularizing a system of polynomial equations. The main computational device is the numerical irreducible decomposition routine powered by the polynomial homotopy continuation methods.