Speaker: Ryan Graham

Title: Block bootstrapping, and other tricks for resampling dependent data

Abstract
Efron's bootstrap was originally developed with applications to i.i.d data
in mind, and did not apply directly to dependent data.  There have been a
number of generalizations of the bootstrap in this direction, one of which
is the block bootstrap. In this talk, I'll describe the block bootstrap
algorithm, a few variations, and present the appropriate theorems that
justify its inclusion within a time-series analyst's standard tool-kit.