Math Seminars

J. Buzzi
(Ecole Polytechnique)

Hyperbolicity from Entropies: some results through inducing schemes


Entropy assumptions ensure some kind of hyperbolicity in low dimension as examplified by the Ruelle-Margulis inequality and Katok's (approximation) theorem on surface diffeomorphisms. For piecewise interval maps, Hofbauer obtained an even stronger (isomorphism) theorem, yielding, e.g., control of the measures of maximum entropy. We previously showed how this strategy could be extended to high dimensional smooth maps assuming only a robust entropy "entropy-expansion" condition. We shall discuss some hyperbolic versions of this approach, both in dimension 2 (arbitrary piecewise affine surface homeomorphisms) and in high dimensions (under an assumption of dominated splitting). We will explain how entropy bounds and inducing give a way around the (explict) building of Markov partitions.