The use of neural networks in γ-π discrimination

A neural network algorithm has been applied to the problem of discriminating between photons and neutral pions using calorimetric information from the ALEPH detector. Results are presented comparing the neural network with the existing ALEPH algorithms. In all cases the performance of the neural network is comparable or superior to that of the conventional algorithms. In particular, at high energies, where these algorithms perform poorly, the neural network is still able to distinguish between the two particle types.