BSB Radar Pulse Classification
P. P. Meiler, A. M. van Wezenbeek · 1990
Abstract : The Brain-State-in-a-Box (BSB) neural network is an auto- associative network. It is able to associate vectors with other vectors. Auto-association results in a system that has been learned to respond strongly to each input vector of a certain set of input vectors by returning this same vector multiplied by its recollection strength, where vectors are bounded by the box. This gives the possibility for convergence to one vector of the learned set of vectors if the actual input vector is not known to the system. When radar signals are translated into vectors, the BSB network can be used to for radar pulse classification. The whole process of classification together with the relation with other non-neural algorithms is shown. Keywords: Netherlands, Translations.