Minutes of HTF on 30/03/00

Yuanning Gao - Update on B tagging at high energies.

Yuanning presented the corrected plots for the boosted sphericity and 
multiplicity for the Z return and semileptonic WW decays presented in the 
previous HTF. The agreement between data and 'smeared' MC looks satisfactory.


Jinwei Wu "Interpolation in NN 4jet"

Jinwei presented an investigation of work carried out within the Wisconsin 
group on interpolating the signal mass, NN output and 
efficiencies (2b and 4b) between different energies using a combination of 
the distance to threshold (as shown by D. Smith previously) and a linear 
interpolation between the energies. He used 199.5 and 201.6GeV MC to 
interpolate to 200GeV and compare to MC made at this energy. Reasonable 
agreement was found for the signal mass and NN output but the efficiencies, 
especially 4b were more problematic.

Some discussion on what MC should be generated for this year followed. It was 
felt than an interpolation over 1-2GeV is tolerable but a larger step may 
start to cause significant differences. The effect of extrapolation was felt 
to be unknown.


Jennifer Kile "A flavour independent Hnunu Neural Net"

Jenn presented the work that she is carrying out on a flavour independent 
analysis; the btag is removed from the NN and placed in the cl calculation.
Provided that a full range of signal masses is used for training then both 
the efficiency and optimisation at 107GeV are as for the usual case for the SM higgs.. A significant improvement in efficiency for H->cc, gg and WW is observed.

She was asked to run through the Cl calculation for all masses for ALEPH week.



Gavin Davies "Improvements in Hnunu cuts analysis"

Gavin presented work that he and Richard are carrying out to try to improve 
Hnunu cuts, in particular with respect to the bb(+isr) background that 
dominates close to threshold. A detailed study of bb MC passing the final 
cuts (supported also by year's data) revealed that most events are 3 jet 
like, sometimes with visible ISR whilst signal is not.
The proposal 
is to force the event into 3 jets. The angular distribution between the jets 
and their electromagnetic fraction are under investigation.  
Cuts based on y23 and the ratio of mvis/mrec were also proposed. Initial 
results indicate that the bb background can be significantly reduced and the 
overall sensitivity of the analysis improved slightly. Final results will be 
shown in ALEPH week.