Application of Bootstrap-Sampling in Random Forest Generating 30 samples for Bootstrap Sampling in 3:2 ratio. Modelling on all 30 samples and finding MSE with respect to each model. Finding mean of 30 samples. Here we dont have the standard deviation of mean_squared_error population so we are going to use second method in which we find the standard deviation of our sample.