1: \begin{abstract}

2: %\scriptsize

3: A scheme for locally adaptive bandwidth selection is proposed which

4: sensitively shrinks the bandwidth of a kernel estimator at lowest

5: density regions such as the support boundary which are unknown to the

6: statistician.

7: In case of a H\"older continuous density, this locally minimax-optimal

8: bandwidth is shown to be smaller than the usual rate, even in case of

9: homogeneous smoothness. Some new type of risk bound with respect to a

10: density-dependent standardized loss of this estimator is established.

11: This bound is fully nonasymptotic and allows to deduce convergence

12: rates at lowest density regions that can be substantially faster than

13: $n^{-1/2}$. It is complemented by a weighted minimax lower bound which

14: splits into two regimes depending on the value of the density. The new

15: estimator adapts into the second regime, and it is shown that

16: simultaneous adaptation into the fastest regime is not possible in

17: principle as long as the H\"older exponent is unknown. Consequences on

18: plug-in rules for support recovery are worked out in detail. In

19: contrast to those with classical density estimators, the plug-in rules

20: based on the new construction are minimax-optimal, up to some

21: logarithmic factor.

22: \end{abstract}