Classification of human cognitive processes by the use of an improved neural backpropagation network

The analysis of human cognitive processes constitutes a great challenge for scientists of all research fields. Better methods in signal processing (that means higher resolution in time and frequency) lead to an even better explanation of the human thinking. An unsolved problem in psychology is the question of whether a change between pictorial and conceptual thinking could be the source of intellectual creativity. We attempt to show results for the classification of such elementary cognitive processes by the use of a neural backpropagation network. In one result, we introduce an advanced backpropagation algorithm, which accelerates the learning speed and stabilizes the learned net through adaptive recursive estimation methods