]notes from HTF meeting (16/03/2000) b-tagging in high energy data -- Yuanning ------------------------------------------ The b-tagging in Z-return events and in semi-leptonic WW events was studdied. The Z-return events were analysed following a QCD analysis.The analysis showed a slight excess of data events compared to MC. Both NN4V and NN6V were checked with and without d0/z0 smearing. A following check was performed on each of the variables used as an input to the NN's. Some excess was observed in the sphericity, but this has a small effect.(Steve commented that new versions of spericity variable have better agreement). A semi-leptonic WW analysis was developed with an approximate efficiency of 70%. A good agrement between data and MC events was observed. Again both NN4V and NN6V were checked, and all variables used in the NN's. In conclusion:- The study supports d0 and z0 smearing. The effects of the smearing on Higgs analysis is small and cannot explain the observed deficite. The b-tagging systematics in Higgs results for winter confs are solid. Cuts combination -- Pedro ----------------------------- The addition of the 189 GeV analyses for the MSSM cuts based selection was presented. The lepton, neutrino and tau final states as well as the hA-4jets analyses were added. The hZ/hA overlap was not treated since the sensitivity to hA(not hZ) was the major motivating factor. Results from addition of 189GeV data. The expected sensitivity to the hZ signal is unchanged at 105.8GeV. The expected sensitivity to the pure hA signal is increased by 1.3 GeV (form 85.6 to 86.9GeV). PAPER -- general discussion. ---------------------------- Members of the HTF discussed plans for the publication of the Standard Model and MSSM Neutral Higgs at 192-202 GeV results. Steve Armstrong will be the editor. The publication will closely follow the winter conference note with minimal modifications and additions. The detector section will be reinserted. About a paragraph of text should be added concerning the BEHOLD! online analysis. The HTF is satisfied with the investigations concerning the deficit of events observed in data; no apparent problem with Monte Carlo or data has been found. Hence statements such as "the deficit is still under investigation" will be removed. The CUTS analysis results of MSSM neutral Higgs bosons will be included when MSSM plane plots are available. Since the Signal Estimator paper has yet to be published, a definition of CL_SE should be inserted and perhaps the value of 95% C.L. limits should be quoted for CL_s+b/CL_b as well as for our standard, CL_SE. We will include median as well as mean expected confidence level curves on C.L. vs. M_H plots. Analysis Improvements. ---------------------- 4-Jets cuts Pairing -- john. ---------------------- The introdustion of the HZ matrix element as a possible variable for pairing choices was presented. Matrix element considering Z decay only. A possible combination of the matrix element and the decay angle pdf was introduced. In conclusion more studies are required. Jets above threshold -- jinwei. ------------------------------- The possibility of using the raw higgs mass instead of the reconstructed mass in the 4 jet NN analysis was studied. This possibility leads to increased sensitivity above threshold. Largest data set this year is unlikely to be at highest energy. Last years energies were used with appropriate weightings to simulate the possible data set from this year (with 200 GeV assumed to be largest data set). Mass definitons for 196 and 200 data were changed above their threshold, and the region around the 202 threshold was checked. In conclusion it was found that the change of the mass definition above threshold led to no improvement. Visible mass in hnn cuts -- richard. ------------------------------------ The possibility of using the visible instead of reco mass in hnn cuts was considered. A comparison was performed at varying b-tags with CL(SE) computed for the visible and reconstructed mass (168pb-1 at 200GeV). The optimal b-tag for both visible and reco mass was found to be 1.05. In conclusion the reco mass was found to perform the best and so will be used this year. Possible improvements to the cuts will be investigated. Hll improvements -- tom. ------------------------ Improvement to the H-tautau channel. By not applying th WW cuts to events with identified tau's the efficiency is raised by approx 5% with no increase in qq or WW bgd. The tau Identidification eff is increased from 54% to 82% by using the charged jet mass (CJM <1.8GeV) instead of the jet mass (JM <2.7GeV). No change in qq or WW bgds is observed. An improved FSR selection for events with 2 identified leptons (only) allows a reduction of qq and WW bgd by approx 30% with little effect on signal. Increaseing efficiency above threshold. Altering mass cut from Mz > 77.5 to Mz > 77.5 or Mz + Mh > 77.5 +threshold lead to new candidates only above threshold. This gives sensitivity above threshold with simple cut, although this new sensitivity has a negligable effect on the C.L. New Channels Study -- jean-michel. ---------------------------------- Jean-michel presented new searches for higgs bosons in the 2DHM framework. h-cc and h-gg were considered in models were the coupling to down-type quarks is suppressed. Assuming h-cc 100%, and then h-gg 100% sucessively. Searching for both decays in the 4jets channel, both analyses share the same pre-selection. Good data MC agreement is seen when using 86.3 pb-1 at 200 GeV. For both analyses a dedicated higgs analysis is developed at each higgs mass in 5 GeV steps. Neural Networks are used to simplify variable management For h-gg a Neural Net using event shapes is developed. For h-cc an anti qq-ZZ(event shape) NN is developed, and an anti WW (pairing +event shape) NN is also developed. A dedicated C-tagging NN is also produced to discriminate C jets from u,d,s jets, which is similar to the existing b-tag in all aspects except that the multiplicity of the eflow ojects/log(E) is used while in the ctag it is the charged multiplicity /log(E). C-tag trained using ZZ events at 189 GeV with 4-jet events of the same flavour. In h-cc the two atin qq and anti WW NN's are merged. Eficciency curves for each analysis are computed for various higgs masses. The working points are determined by minimising the Confidence level based on counting only with 160pb-1. The optimisation is performed for higgs mass hypotheses in 5 GeV bins (corresponding to the produced MC). The efficiency given by each analysis (optimised at a specific mass), for all masses is then computed. Then for a given mass hypothesis MH, the efficiency of the analysis at this mass is computed as a simple linear interpolation from the two nearest optimisation masses. The expected confidence levels are 71.2 GeV and 75.1 GeV for the h-cc and h-gg analyses respectively. Future work includes improving the analyses, the C-tag, and incorporating the hll and hnn, a full scan of the 2HDM is also planned. In conclusion the analyses for h-cc and h-gg already exist and there was a request for them to be added to the online-analysis. The next presentation of these channels is set for the comming aleph week Additions to online-analysis -- general discussion. --------------------------------------------------- It has been noted that analyses that are not covered by the current online package could possibly be added but this would greatly increaese the amount of work required for steve + jason. One possible solution was to offer people responsible for their analyses, (H+H-,2HDM) the chance to produce their own online pages which could be accessed via a link from the BEHOLD page. It is hoped that this possibility will be explored more. (aspects of security are still to be considered).