1: \begin{abstract}                          % Abstract of not more than 200 words.

2: Many problems arising in control

3: require the determination of a mathematical model of the application.

4: This has often to be performed starting from input-output data,

5: leading to a task known as system identification in the engineering literature.

6: One emerging topic in this field is estimation of networks consisting of several interconnected

7: dynamic systems.

8: We consider the linear setting assuming that system outputs are the result of many correlated inputs,

9: hence making system identification severely ill-conditioned.

10: This is a scenario often encountered when modeling complex systems composed by many sub-units with feedback

11: and algebraic loops.

12: We develop a strategy cast in a Bayesian regularization framework

13: where any impulse response is seen as realization of a zero-mean Gaussian process.

14: Any covariance is defined by the so called stable spline kernel which includes

15: information on smooth exponential decay.

16: We design a novel Markov chain Monte Carlo scheme able to reconstruct the

17: impulse responses posterior by efficiently dealing with collinearity.

18: Our scheme relies on a variation of the Gibbs sampling

19: technique: beyond considering blocks forming a partition of the parameter space,

20: some other (overlapping) blocks are

21: also updated on the basis of the level of collinearity of the

22: system inputs. Theoretical properties of the algorithm are

23: studied obtaining its convergence rate. Numerical experiments

24: are included using systems containing hundreds of impulse

25: responses and highly correlated inputs.

26: \end{abstract}

27: