February 8th, 2008, 12:41 pm
Hi all, I am trying to understand Kullback-Leibler or Jensen-Shannon divergence between two distributions. I searched on net for that, but most the cases, inclusing Wikipedia, give only definition of that. However I am not getting the intuition or idea behind that. My question is why that expression? How it measures the distance? If people here share their idea, understanding [in little bit easy language] on that, or possible e-books etc it will be very helpful for me.Regards,