CONFORMAL MEASURES AND DENSITY ESTIMATION
Manfred Denker
Tina Facca
Abstract: The notion of conformal measures or densities goes back to Patterson’s work [8]. By a
theorem of Milnor and Thurston [7] a piecewise monotone and continuous map of the
interval is semiconjugate to one with constant slopes. The semiconjugacies can be defined
by distribution functions of conformal measures as shown in [2]. In this note we
show that for some transformations the conjugacies are estimable functions and
can be used to improve estimation procedures, in particular density estimations.
2000 AMS Mathematics Subject Classification: Primary: 62G05, 37E05; Secondary:
62G07, 62G09, 37A05, 37A50.
Keywords and phrases: Resampling, interval map, conjugacy, density estimation.