In this paper, we develop a methodology to estimate hidden linear networks when only an aggregate outcome is observed. The aggregate observable variable is a linear mixture of the different networks and it is assumed that each network corresponds to the transmission mechanism of different...