Stability and relaxation time of Tank and Hopfield's neural network for solving LSE problems

A.D. Culhane et al. (1989) proposed a fast technique for computing discrete Hartley and Fourier transforms using a Tank and Hopfield linear programming neural network. It was proved mathematically that the network is stable under some conditions. The network can also be used for solving linear least squares error (LSE) problems. It is shown that the stability of the network is guaranteed even under weaker conditions. The author also provides a more accurate solution for the network relaxation constants and discusses the accuracy of computation.<
>