Introducing resampling methods
Resampling methods are a set of techniques based on the use of subsets of data, which can be extracted either randomly or according to a systematic procedure. The purpose of this technology is to approximate some characteristics of the sample distribution – a statistic, a test, or an estimator – to validate a statistical model.
Resampling methods are one of the most interesting inferential applications of stochastic simulations and the generation of random numbers. These methods became widespread during the 1960s, originating from the basic concepts of Monte Carlo methods. The development of Monte Carlo methods took place mainly in the 1980s, following the progress of information technology and the increase in the power of computers. Their usefulness is linked to the development of non-parametric methods, in situations where the methods of classical inference cannot be correctly applied.
The following details can be observed from...