Hi, may I check for autoencoder training (AutoEncoder.m) :
[U, CList] = TrainNetwork(in, out, U, M, lambda, iter);
Why the out used is the noisy data? Isn't the output of the network supposed to be compared against noiseless data (tout in this case)?
Second question is, according to this paper https://arxiv.org/abs/1902.10445, the unitary of the entire network should have dimension of 2^(m+n), where m and n are the number of qubits of input and output, but the U here only has the dimension of 2^m. Why is that so?
Thanks for replying.
Hi, may I check for autoencoder training (AutoEncoder.m) :
[U, CList] = TrainNetwork(in, out, U, M, lambda, iter);
Why the out used is the noisy data? Isn't the output of the network supposed to be compared against noiseless data (tout in this case)?
Second question is, according to this paper https://arxiv.org/abs/1902.10445, the unitary of the entire network should have dimension of 2^(m+n), where m and n are the number of qubits of input and output, but the U here only has the dimension of 2^m. Why is that so?
Thanks for replying.