1: \begin{abstract}

2: 		For $D$ a bounded domain in $\mathbb R^d, d \ge 2,$ with smooth boundary $\partial D$, the non-linear inverse problem of recovering the unknown conductivity $\gamma$ determining solutions $u=u_{\gamma, f}$ of the partial differential equation 

3: 		\begin{equation*}

4: 		\begin{split}

5: 		\Div(\gamma \Grad u)&=0 \quad \text{ in }D, \\

6: 		u&=f \quad \text { on } \partial D,

7: 		\end{split}

8: 		\end{equation*}

9: 		from noisy observations $Y$ of the Dirichlet-to-Neumann map \[f \mapsto \Lambda_\gamma(f) =  {\gamma  \pd{u_{\gamma,f}}{\nu}}\Big|_{\partial D},\] with $\partial/\partial \nu$ denoting the outward normal derivative, is considered. The data $Y$ consists of $\Lambda_\gamma$ corrupted by additive Gaussian noise at noise level $\eps>0$, and a statistical algorithm $\hat \gamma(Y)$ is constructed which is shown to recover $\gamma$ in supremum-norm loss at a statistical convergence rate of the order $\log(1/\eps)^{-\delta}$ as $\eps \to 0$. It is further shown that this convergence rate is optimal, up to the precise value of the exponent $\delta>0$, in an information theoretic sense. The estimator $\hat \gamma(Y)$ has a Bayesian interpretation in terms of the posterior mean of a suitable Gaussian process prior and can be computed by MCMC methods.

10: 		

11: 		\smallskip

12: 		

13: 		\noindent\textit{Keywords: non-linear inverse problems, elliptic partial differential equations, electrical impedance tomography, asymptotics of nonparametric Bayes procedures}

14: 	\end{abstract}

15: