What Next? cont’d
Make improvements to algorithm as measured by #MC particles per cluster, probably in two parts:
- when algorithm encounters cell touching 2 pre-existing clusters, it always merges them. It has sufficient information to guess when this may be due to overlapping showers or jets and perhaps handle the situation better.
- need to add a real refinement step after the initial crude clustering is done which attempts to use more sophisticated pattern recognition on the characteristic forms of energy deposition of different particles.
- Associate clusters and tracks (outside scope of this discussion).