Scattering network used for digit recognition consisting of two or three fully connected layers with N 1 = 128, N 3 = 10 and either without a hidden layer or with N 2 {20, 30, 60, 80}. We consider equal decay rates , set the intrinsic decay to zero ( '= 0) at the probe sites and start from J / = 2 with an added random disorder. The input consisting of 64 grey-scale pixel values is encoded in the detuning of the first layer to which we initially add a trainable offset. A vector of pixel values serves as the input. We choose to detune the background to x j = 5 and make the foregroundthe numeralsresonant, that is, x j = 0. The inset illustrates the nonlinear effect of the first layer, showing the real and imaginary parts of ({[G 1(0)]} j, j) (equation (13)). The response to a probe signal at the third layer constitutes the output vector. The index of maximal y = Im S , constitutes the class.
