Conjugate gradient algorithms for complex-valued neural networks

Abstract

In this paper, conjugate gradient algorithms for complex-valued feedforward neural networks are proposed. Since these algorithms yielded better training results for the real-valued case, an extension to the complex-valued case is a natural option to enhance the performance of the complex backpropagation algorithm. The full deduction of the classical variants of the conjugate gradient algorithm is presented, and the resulting training methods are exemplified on synthetic and real-world applications. The experimental results show a significant improvement over the complex gradient descent algorithm.