Nonuniformly expanding 1D maps (Q852407): Difference between revisions

Let I be a closed interval or a circle, and let \(C^2(I,I)\) denote the set of \(C^2\) maps from I to itself. Firstly, the authors define a class \({\mathcal M}\subset C^2(I,I)\) such that the maps in \({\mathcal M}\) are among the simplest with nonuniform expansion. Then, in a neighborhood of each \(f_0\in{\mathcal M}\), an admissible set of perturbations \({\mathcal G}(f_0)\) is introduced. Finally, the set of ``good maps'' \({\mathcal G}\) is defined as \({\mathcal G}=\cup_{f_0\in{\mathcal M}}{\mathcal G}(f_0)\). For this set \({\mathcal G}\), the authors prove that every \(f\in{\mathcal G}\) has an absolutely continuous invariant probability measure. A further result, concerning the 1-parameter families of 1D maps \(\{f_a\in C^2(I,I),a\in (a_1,a_2) \}\), is the following: For every \(C^2\) family \(\{f_a\}\) satisfying some ``parameter transversality condition'' at \(f_0\in{\mathcal M}\), the set \(\{a: f_a\in {\mathcal G}\}\) has positive Lebesgue measure.