Sums of squares of polynomials and graphs
We investigate a hierarchy of semidefinite bounds for the stability number α(G) of a graph G, based on its continuous copositive formulation:
α(G)=min{t:t(I+AG)−J∈COPn},
where n=|VG|, AG is the adjacency matrix of G and J is the all-ones matrix. Here, COPn is the copositive cone, consisting of the symmetric n×n matrices M for which the polynomial pM=∑ni,j=1Mijx2ix2j is nonnegative over Rn. By replacing the copositive cone COPn by its inner approximation, consisting of the matrices M for which the polynomial (∑ix2i)rpM is a sum of squares, we obtain the bounds ϑ(r)(G), known to converge asymptotically to α(G). De Klerk and Pasechnik (2002) conjectured that ϑ(r)(G)=α(G) for any r≥α(G)−1, but even the weaker conjecture asking whether equality holds for some r is still open. We discuss old and new results around these questions, which display a nice interplay between graph structure, polynomial optimization and real algebraic geometry.
Joint work with Luis Felipe Vargas (CWI, Amsterdam).