Minisymposium on Algebraic Geometry and Optimization

SIAM Conference, Darmstadt, May 16-19, 2011

ABSTRACT
The interplay of semidefinite programming and algebraic geometry has
recently led to new paradigms for solving or approximating hard
optimization problems. Many nonlinear optimization problems can be encoded
into semidefinite programs through the theory of moments and positive
polynomials. The basic idea underlying this approach is that, while
testing whether a polynomial is nonnegative is a hard problem, the relaxed
problem of testing whether it can be written as a sum of squares of
polynomials can be reformulated as a semidefinite program. This symposium
aims to highlight the various algebraic tools and methodologies behind
this approach.