1: \begin{abstract}

2: We study the error in approximating the minimum of a Brownian motion

3: on the unit interval based on finitely many point evaluations.

4: We construct an algorithm that adaptively chooses

5: the points at which to evaluate the Brownian path.

6: In contrast to the $1/2$ convergence rate of optimal nonadaptive algorithms,

7: the proposed adaptive algorithm converges at an arbitrarily high polynomial rate.

8: \end{abstract}

9: