A conditional game for comparing approximations


Frederik Eaton ;
Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics, PMLR 15:63-71, 2011.


We present a “conditional game” to be played between two approximate inference algorithms. We prove that exact inference is an optimal strategy and demonstrate how the game can be used to estimate the relative accuracy of two different approximations in the absence of exact marginals.

Related Material