Drift chamber tracking with neural networks

Abstract

Drift chamber tracking with a commercial analog VLSI neural network chip is considered. Voltages proportional to the drift times in a four-layer drift chamber were presented to the Intel ETANN (Electrically Trained Analog Neural Network) chip. The network was trained to provide the intercept and slope of straight tracks traversing the chamber. The outputs were recorded and later compared off-line to conventional track fits. Two types of network architectures were studied.<>