1: \begin{abstract}

2: Stochastic approximation is a framework unifying many random iterative

3: algorithms occurring in a diverse range of applications.  The stability of

4: the process is often difficult to verify in practical applications and the

5: process may even be unstable without additional stabilisation techniques. We

6: study a stochastic approximation procedure with expanding projections

7: similar to Andradóttir [\textit{Oper.~Res.} 43 (2010) 1037--1048]. We focus

8: on Markovian noise and show the stability and convergence under general

9: conditions.  Our framework also incorporates the possibility to use a random

10: step size sequence, which allows us to consider settings with a non-smooth

11: family of Markov kernels. We apply the theory to stochastic 

12: approximation expectation

13: maximisation with particle independent Metropolis-Hastings sampling.

14: \end{abstract}

15: