Efficient boostrap methods: a review

One of the fundamental of mathematical statistics is the estimation of sampling characteristics of a random variable, a problem that is increasingly solved using bootstrap methods. Often these involve Monte Carlo simulation, but they may be costly and time-consuming in certain problems. Various methods for reducing the simulation cost in bootstrapp simulations have been proposed, most of them applicable to simple random samples. Here we review the literature on efficient resampling methods, make comparisons, try to assess the best method for a particular problem.