Wild Cantor attractors exist (Q1909390)

It has been known for a while that a smooth unimodal map of an interval either has an attracting periodic orbit, a transitive cycle of intervals, or a solenoidal attractor. The question posed by Milnor was whether there might be sets with basins of attraction of positive measure coexisting with the topologically transitive situation. This question was an important challenge for the theory in the late eighties and early nineties. By now, we realize that the difficulty of the problem is in its dependence on the degree of degeneracy of the critical point of the map. It is known, for example, that the answer to Milnor's question is negative for the logistic family. As the present paper shows, an example can be found when one considers maps with a sufficiently degenerate critical point. The paper provides a very poor estimate of what the order of the critical point should be. The way in which the positive measure is obtained is quite new in one-dimensional dynamics. The basin of the attractor is never ``constructed'' in a geometrically apparent way. Instead, the positive measure is proved by a probabilistic argument. As a result, the geometric estimates getting into the proof need not be very strong. It should be noted that this probabilistic method was more recently used by Nowicki and Van Strien to show a polynomial with the Julia set of positive measure.