Neurocomputing methods for pattern recognition in nuclear physics-Elastic tracking

The recent progress on the development and application of novel neurocomputing techniques for pattern recognition problems of relevance to relativistic heavy ion collider experiments is reviewed. The elastic tracking (ET) algorithm is shown to achieve sub-pad two track resolution without preprocessing. ET can extend tracking capabilities to much higher densities than possible via conventional road finding or even previously proposed novel Hopfield network algorithms. ET is an adaptive template matching algorithm formulated in terms of dynamical systems