How to determine the strength of the mutual information value

Asked Jun 28 '21 at 15:29

Active Jun 28 '21 at 16:13

Viewed 50 times

I am reading about mutual information dependency measures. I found that it ranges from $0$ to infinity.

I have one simple question. How can I measure the strength of the dependency using this method? For example, Kendall's $\tau$ ranges from $-1$ to $1$, and we can simply understand the strength of the dependency. What is the corresponding interpretation of mutual information?

edited Jun 28 '21 at 16:13

Arya McCarthy

8,787

asked Jun 28 '21 at 15:29

Maryam

1,620

2

You may be interested in a question of mine from last year on what I call "information correlation". That "information correlation" equation I give seems to have something to it when I have run simulations (if I have four independent variables and calculate the information correlation between the full four ($X$) and a subset of three ($Y$), I get $\rho_I(X, Y) \approx 0.75$, as I would expect). – Dave Jun 28 '21 at 15:40

How to determine the strength of the mutual information value

0 Answers0