Inequalities for Multivariate Polynomials

We summarize researches – in great deal jointly with my host Y. Sarantopoulos and his PhD students V. Anagnostopoulos and A. Pappas – started by a Marie Curie fellowship in 2001 and is still continuing. The project was to study multivariate polynomial inequalities. In the course of work we realized the role of the “generalized Minkowski functional”, to which we devoted a throughout survey. Resulting from this, infinite dimensional extensions of Chebyshev’s extremal problems were tackled successfully. Investigating Bernstein-Markov constants for homogeneous polynomials of real normed spaces led us to the application of potential theory. Also we found at first unexpected connections of polarization constants of R and C to Chebyshev constants of S and S, respectively. In the study of polarization constants, a further application of potential theory occurred. This led us to realize that the theory of rendezvous numbers can be much better explained by potential theory, too. Our methods for obtaining Bernstein type pointwise gradient estimates for polynomials were compared in a recent case study to the yields of pluripotential theoretic methods. The findings were that the two rather different methods give exactly the same results, but the two currently standing conjectures mutually exclude each other. 1 Polynomials In Higher Dimensions In the whole paper K stands for either R or C. If X is a normed space, X = L(X,K) is the usual dual space, and S := SX , S := SX∗ , B := BX and B := BX∗ are the unit spheres and (closed) unit balls of X and X , respectively. Moreover, P = P(X) and Pn = Pn(X) will denote the space of continuous (i.e., bounded) polynomials of free degree and of degree ≤ n, resp., from X to K. There are several ways to introduce continuous polynomials over X , one being the linear algebraic way of writing Pn := P 0 + P 1 + · · ·+ P n , and P := ∞ ⋃