0

I want to compare the mutual information for each possible pair of variables. I want to identify the lower and higher mutual information of all possible combination. I'm not sure if doing the traditional way is correct. Maybe some kind of normalization would be appropriate for relative comparisons.

My variables are continuous in the range 0-1 with unknown distribution.

I also look for packages for mutual information like sklearnon

Brayan T
  • 1
  • 1
  • What kind of random variables are you working with (discrete or continuous)? Maybe more generally, do you have any extra knowledge about your problem (what space you are looking at, do the random variables follow some particular distributions, ...)? – adrien_vdb May 16 '23 at 05:08

0 Answers0