Central Limit Theorems for Conditional Markov Chains


Mathieu Sinn, Bei Chen ;
Proceedings of the Sixteenth International Conference on Artificial Intelligence and Statistics, PMLR 31:554-562, 2013.


This paper studies Central Limit Theorems for real-valued functionals of Conditional Markov Chains. Using a classical result by Dobrushin (1956) for non-stationary Markov chains, a conditional Central Limit Theorem for fixed sequences of observations is established. The asymptotic variance can be estimated by resampling the latent states conditional on the observations. If the conditional means themselves are asymptotically normally distributed, an unconditional Central Limit Theorem can be obtained. The methodology is used to construct a statistical hypothesis test which is applied to synthetically generated environmental data.

Related Material