Neural network based neutral particles reconstruction with the E687 hadron calorimeter

We present a neutral particle reconstruction algorithm based on a neural network approach applied to the E687 hadron calorimeter. A measurement of the invariant mass of the Σ± → nπ± is presented to verify the reliability of the reconstruction. The reconstructed invariant mass of the charmed meson D+ → KL0π+π+π− is also presented to show the possible application of this technique to charmed particles decaying into a neutral hadron. An example of this would be Λc+ → nK−π+π+.