What metric best captures the information loss when binning a large data-set into a histogram representation?

Suppose I am looking at millions of data sets each with millions of data points and I need to capture details about each of those distributions with as much accuracy as possible. Histograms are a...

07 July 2013 8,152 7 View