Sampling with & without replacement

[csawtelle]

Sampling with replacement, calculate the parameter. Sampling with replacement just means that when you "take out" the sample, you leave a duplicate in its place.

Ex Sample without replacement:
data = {1,2,3,4,5} sample out 4 - result data = {1,2,3,5}
another sample
data = {1,2,3,5} sample out 1 - result data = {2,3,5}
etc...

Sampling with replacement just means that you can sample out the same number multiple times, since you leave the "duplicate" inside the dataset.

[csawtelle]

http://vda.univie.ac.at/Teaching/ML/15s/LectureNotes/11_sampling.pdf

[csawtelle]

https://www.youtube.com/watch?v=7oWbmME2R_k