Finding complex roots of polynomials by feedforward neural networks

This paper extended the finding real roots of polynomials by feedforward neural nehvorks (FNN) with constrained learning algorithm (CLA) to finding complex roots of arbitrary polynomials (including complex ones). Likewise, for those high order complex polynomials, this network model is also extended to one which fin& recursively a small number of complex roots af a polynomial (less than the total number of roots lo be found) at a time. Finalb some experimental results are presented.