F.P. Harter, H.F. de Campos Velho (2002): Elman and Jordan Neural Networks for Solving XOR Problem, WorkCAP-2002: Workshop of the Applied Computing Courses of the INPE, 20-21 November, Sao Jose dos Campos (SP), Brasil.

Abstract: The Elman and Jordan's artificial neural networks are used for solving the XOR problem. This problem has been employed to test the performance of neural networks, since the XOR problem can not be solved by the linear perceptron. The Elman and Jordan's are particularly important, because they present recurrent connections. From the success of these neural networks for solving teh XOR problem, it has motivated the authors to apply them to neurodynamics problems, such as the data assimilation.