X complete year 2 report
X run any job using PAEDM
lepton tools
------------
reproduce e and mu isolation work using PAEDM - may need full MC
simulation. see if using Z decays or W decays in other final states
allows rapid progress. Are all tracks/clusters stored in
ESD?
understand (with Paganis?) number of muons per event, double
counting issues, b quarks as a source of muons. Are all
tracks/clusters stored in ESD?
ditto for electronsAre all tracks/clusters stored in
ESD?
X study ET dis for leptons from W decay in ttH.
X Agree with experts where to set cut for generating
X agree how many full MC to be generated in short-term and long-term
X for DC2 and at what higgs masses. understand why these higgs masses
X have been chosen and waht is useful mass range for us to cover.
work with experts to verify that ttH samples generated are correct
X at truth level (check for silly bugs) work in progress
-
waiting for BM to reply 12.10.2004: Then investigate truth (=HepMC cut-down?)
act as customer for muon and electron suplliers
use best energy corrections. verify how ET(reco)/EET(truth)
varies for electrons and muons with eta. -- understand treatment
of photons radiated from electrons and 3x7 clustering v 5x5 clustering
3x7, 5x5 see TDR. Magnet bends charged trajectories.photon tangential so
elongated is better.
Kyle: e- ok in PAEDM - egammarec starts with clusters and gets tracks.
Softebuilder starts with tracks and gets clusters.
DoElectron seems to have arbitary cuts.
mu's - problem: apparently no correction for missing Et - off due to detector
losses.
b-jet finding
-------------
learn from heavy flavour people best methods for finding b-jets
(a) leptons (b) vertex tagging (c) jet shape??
understand use of these methods for multi-b final state
maybe presentations from btagging in SW week (sep20-24) might help.
Also vadimir's talks.
Package possibly in development.
t tbar finding
--------------
understand method of reconstructing top -->bW, W-> 2 jets
and top --> bW, W--> lnu
learn from top search people and copy their methods in detail
learn about b-tagging in top search using (a) leptons from b-decay
(b) vertex info from b-decay
apply these methods to ttH and ttjj samples.(also ttZ??)
learn about backgrounds to ttjj, from 8j final state (??)
someone will be simulating these.
apply these methods at truth level (getting v narrow mass peaks)
understand why efficiency is not 1 (all the loss factors with
a perfect detector)
apply these methods at truth level to backgrounds to t tbar.
understand how well t tbar signal is found above non ttbar background
(a) in ttbar final state (b) in ttbarjj state when m(jj) is high
b bbar jet pairs
----------------
assume truth level and no combinatoric problems. reconstruct
higgs mass peak for h-->b bar. understand biases, width, efficiency
losses due to ET and eta cuts
do the same at reco level to find m(bbbar) assuming correct b-jets
have been identified. understnad biases, width, efficiency losses
due to jet finding and measurement errors.
consider whether these biases can be overcome
consider whether some kinematic cuts will improve sharpness of signal
now use realistic b-tagging algorithm and repeat
understand efficiencies
t tbar b bbar events
-------------------
repeat b bbar analysis above using events with t tbar tag
obtain higgs mass peak for ttH, H-->bb final state
repeat analysis on background channels to obtain backgrounds
vary cuts and procedures to optimise significance of discovery.
write up results
obtain ph d