Hi all, I'm wondering how to get some value like P-value from D_KL? the smaller the D_KL, the more similar the two distributions, but how small of this value could proof there is no significant difference between them? Could we...

Kullback–Leibler divergence

Topics Kullback–Leibler divergence Forum Thread

Posts 1 - 1 of 1

duxg

Hi all,

I'm wondering how to get some value like P-value from D_KL? the smaller the D_KL, the more similar the two distributions, but how small of this value could proof there is no significant difference between them? Could we also get something like P-value?

Thanks

all the best,
Xiaoguang

Save

Cancel