Discrete Dynamics in Nature and Society · 2009 · 22 citations · 12 references
ComputingComputer ScienceArtificial IntelligenceMachine LearningOptimizationNeural NetworksPattern RecognitionBackpropagationGradient DescentMultilayer PerceptronNeural Network ModelingStochastic Gradient Descent
The batch split‐complex backpropagation (BSCBP) algorithm for training complex‐valued neural networks is considered. For constant learning rate, it is proved that the error function of BSCBP algorithm is monotone during the training iteration process, and the gradient of the error function tends to zero. By adding a moderate condition, the weights sequence itself is also proved to be convergent. A numerical example is given to support the theoretical analysis.
12
An Extension of the Back-Propagation Algorithm to Complex Numbers
Tohru Nitta · Neural Networks · 1997 · 307 citations