Abstract: We propose a neural operator combining Deep operator network and Fourier neural operator for gain modeling of EDFAs. The operator shows low prediction and generalization errors of 0.0472dB ...