Comment on ´Neural network for estimation of parameters of sinewave´ (and reply)

For the original article see ibid., vol.26, no.11, p.689-91 (1990). In the aforementioned paper by S. Osowski it is claimed that ´with time constants of integrators equal to 10^-6 the correct solution was obtained irrespective of the initial conditions . . . The commenters show that this claim is not true in the general case and that the results then depend on the initial conditions. In reply the author points out that he did not mention that the solution presented in his article leads to the global minimum of the objective function, however, he admits to an inaccuracy in the wording and corrects this.