Discrete Fourier transform using neural network
M. Romero-Tetzicatl, Ruben Alejos-Palomares, R. Castellanos-Nolasco, Felipe Gomez-Castaneda · 2002
In this paper we present a basic scheme which computes the Discrete Hartley Transform (DHT) and Discrete Fourier Transform (DFT) based on a low complexity neural net. A generalization of this work in order to develop other types of discrete transforms is feasible if they can be represented in vectorial form and their transformation matrix is invertible to assure stability of the network.