Electromagnetic Calorimeter Intercalibration with Neural Networks for the CMS Experiment at CERN

The Electromagnetic Calorimeter (ECAL) of the Compact Muon Solenoid (CMS) detector is made of $\sim $76000 lead tungstate (PbWO$_{4})$ crystals. The calorimeter has excellent energy resolution and the small crystal size gives very good position resolution. During the summer of 2006 a production ECAL module was tested jointly with the Hadron Calorimeter in the CERN test beam in the energy range of a few GeV to 300GeV for pions and up to 150GeV for electrons. Given the large number of crystals intercalibration of ECAL is a very important and complicated task. I will discuss different approaches to intercalibration of ECAL including the method of using neural networks.