Finite-Sample Analysis of Fixed-k Nearest Neighbor Density Functional Estimators

Estimating entropies and divergences of probability distributions in a consistent manner is of importance in a number of problems in machine learning.