1: \begin{abstract}

2:   This paper introduces a new method to tackle the issue

3:   of the almost sure convergence of stochastic approximation

4:   algorithms defined from a differential inclusion.  Under the

5:   assumption of slowly decaying step-sizes, we establish that the set

6:   of essential accumulation points of the iterates belongs to the

7:   Birkhoff center associated with the differential inclusion.  Unlike

8:   previous works, our results do not rely on the notion of asymptotic

9:   pseudotrajectories introduced by Bena\"im--Hofbauer--Sorin, which is the

10:   predominant technique to address the convergence problem. They follow

11:   as a consequence of Young's superposition principle for closed

12:   measures.  This perspective bridges the gap between Young's principle and

13:   the notion of invariant measure of set-valued dynamical systems

14:   introduced by Faure and Roth.

15:   Also, the proposed method allows to obtain

16:   sufficient conditions under which the velocities

17:   locally compensate around any essential accumulation point.

18: \end{abstract}

19: