1: \begin{abstract}%

2: This paper studies fixed step-size stochastic approximation (SA) schemes,

3: including stochastic gradient schemes, in a Riemannian framework. It is

4: motivated by several applications, where geodesics can be computed

5: explicitly, and their use accelerates crude Euclidean methods. A fixed

6: step-size scheme defines a family of time-homogeneous Markov chains, 

7: parametrized by the step-size. Here, using this formulation, non-asymptotic

8: performance bounds are derived, under Lyapunov conditions. Then,

9: for any step-size, the corresponding Markov chain is proved to admit a unique stationary

10: distribution, and to be geometrically ergodic.  This result gives rise

11: to a family of stationary distributions indexed by the step-size, which is further shown to 

12: converge to a Dirac measure, concentrated at the solution of the problem at hand, as 

13: the step-size goes to $0$.  Finally, the asymptotic rate of this convergence is established,

14:   through an asymptotic expansion of the bias, and a central limit theorem. %Our results are illustrated by ...

15: \end{abstract}

16: