A first look at different flavours in SGV

1
A first look at different flavours in SGV

• vertex package will contain flavour tag based on
  variables obtained from ZVTOP
• and using the Bristol neural net
• ? need to compare tagging performance with
  that achieved previously
• begin by looking at most
• relevant plots quark flavour
• content, MPt and
• vertex momentum vs MPt
• for u,d,s jets, c and b jets
• at 3 different jet energies
• 45 GeV, 100 GeV, 250 GeV
• right SLD results for comparison

2
45 GeV jets
3
100 GeV jets
4
250 GeV jets
5
Results

• at 45 GeV reasonable agreement with SLD results
  
• better MPt resolution due to improve detector
  (and possibly due to ISR being switched off)
• as one goes up in energy, fraction of b-jets
  before cuts decreases rapidly,
• but because of B decay chain being more
  spread out at higher energy, it can be better
• reconstructed ? ZVTOP finds secondary
  vertices for a larger fraction of the b jets at
• higher energy
• will be looked at in more detail
• if secondary vertex found, separation between b
  and c using MPt only becomes better
• when increasing jet energy
• since vertex momentum increases with increasing
  jet energy, correlation between
• vertex momentum and MPt changes
• ? might need to normalise vertex momentum
  before feeding it into NN