1: \begin{abstract}

2: 

3: Estimation procedures based on recursive algorithms are interesting and powerful techniques that are able to deal rapidly with (very) large samples of high dimensional data. The collected data may be contaminated by noise so that robust location indicators, such as the geometric median, may be preferred to the mean. In this context, an estimator of the geometric median based on a  fast and efficient averaged non linear stochastic gradient algorithm has been developed by \cite{HC}. This work aims at studying more precisely the non asymptotic behavior of this algorithm by giving non asymptotic confidence balls. This new result is based on the derivation of improved $L^2$ rates of convergence as well as an exponential inequality for the martingale terms of the recursive non linear Robbins-Monro algorithm. 

4:      

5: %With the development of automatic sensors, it is more and more usual to deal with large samples taking values in high dimensional spaces such as functional spaces. Means is the most frequently used estimator of localisation. Nevertheless, it is very sensitive to outliers. In this framework, we will focus on the geometric median also called $L^{1}$-median. Many estimation algorithms exist (\cite{Ger08}, \cite{VZ00}, ...) but they are often not fast enough, or they need many computational efforts, or they cannot simply treat new data. We will focus on the stochastic gradient algorithms introduced by \cite{HC}. Many results about their convergence were given, such as the almost sure convergence and the convergence in law. At first, we will give the rate of convergence in quadratic means and prove that this rate is optimal. Secondly, we will use some exponential inequalities to give asymptotic confidence intervals.

6: \end{abstract}

7: