1

I have a set of boolean data where I'm trying to distinguish between data where the values are uniformly distributed and data where the same class is grouped together.

e.g 1001010011 seems pretty uniformly distributed when compared with 1110000011.

What sort of metric can I use to measure how uniformly distributed/grouped the values are?

Quicksilver
  • 11
  • 1
  • Welcome to Cross Validated! I like this question, +1. Some buzzwords (phrases?) that might help you are “binomial autocorrelation” and “Bernoulli autocorrelation”. // Of possible interest: https://stats.stackexchange.com/questions/20563/is-there-a-test-for-independence-in-a-bernoulli-process. (I found that searching for my buzzwords!) – Dave Aug 13 '21 at 00:08
  • I found what I was looking for in https://stats.stackexchange.com/a/20577/332274 – Quicksilver Aug 16 '21 at 16:58

0 Answers0