Chained Kullback-Leibler Divergences.

We define and characterize the "chained" Kullback-Leibler divergence min (‖) + (‖) minimized over all intermediate distributions and the analogous -fold chained K-L divergence min (‖) + … + (‖) + (‖) minimized over the entire path (,…,). This quantity arises in a large deviations analysis of a Markov chain on the set of types - the Wright-Fisher model of neutral genetic drift: a population with allele distribution produces offspring with allele distribution , which then produce offspring with allele distribution , and so on. The chained divergences enjoy some of the same properties as the K-L divergence (like joint convexity in the arguments) and appear in -step versions of some of the same settings as the K-L divergence (like information projections and a conditional limit theorem). We further characterize the optimal -step "path" of distributions appearing in the definition and apply our findings in a large deviations analysis of the Wright-Fisher process. We make a connection to information geometry via the previously studied continuum limit, where the number of steps tends to infinity, and the limiting path is a geodesic in the Fisher information metric. Finally, we offer a thermodynamic interpretation of the chained divergence (as the rate of operation of an appropriately defined Maxwell's demon) and we state some natural extensions and applications (a -step mutual information and -step maximum likelihood inference). We release code for computing the objects we study.