How block bootstrap is different from normal bootstrap?



Recently I studying about bootstrap resampling.

bootstrap- The bootstrap is a flexible and powerful statistical tool that can be used to quantify the uncertainty associated with a given estimator or statistical learning method.

block bootstrap - It is the most general method to improve the accuracy of bootstrap of time series data.

I want to know difference between them and condition of their uses.


Hi @ankit81195,

When sequences are concerned such as time series data, using normal bootstrap method fails because of correlation. So to use bootstrapping on time series data, you would have to consider “blocks” or chunks of data sequences for the sampled data to behave similar to population.

Please go through the wiki for a deeper overview.