1: \begin{abstract}

2: We prove a strong law of large numbers for a class of strongly mixing

3: processes.

4: %The 

5: Our

6: result rests on

7: recent advances in understanding of concentration of measure.

8: It is simple to apply and gives finite-sample (as

9: opposed to asymptotic) bounds, with readily computable

10: rate constants.  In particular, this makes it

11: suitable for analysis of inhomogeneous Markov processes.

12: We demonstrate how it can be applied to establish an

13: almost-sure convergence result

14: for a class of models that includes

15: as a special case a class of adaptive Markov chain Monte Carlo

16: algorithms.

17: \end{abstract}

18: