A geometrical approach to finding multivariate approximate LCMs and GCDs

In this article we present a new approach to compute an approximate least common multiple (LCM) and an approximate greatest common divisor (GCD) of two multivariate polynomials. This approach uses the geometrical notion of principal angles whereas the main computational tools are the Implicitly Restarted Arnoldi method and sparse QR decomposition. Upper and lower bounds are derived for the largest and smallest singular values of the highly structured Macaulay matrix. This leads to an upper bound on its condition number and an upper bound on the 2-norm of the product of two multivariate polynomials. Numerical examples are provided.