From narain@buphy.bu.edu Wed Sep 20 10:58:37 2006 Date: Wed, 20 Sep 2006 09:58:37 -0400 (EDT) From: Meenakshi Narain To: Sarah Catherine Eno Cc: Sridhara Dasu , Meenakshi Narain Subject: btagging proposal for support from USCMS Dear Sarah, and Sridhara, Below is the request for endorsement and support from US CMS "management" for my proposed activities on CMS. Our DOE site visit (at Brown) is next week and it would be nice to get your feedback, even if informally, so that I can relay your (strong :)) endorsement to Saul Gonzales. He is a "process" person and if were to go by our previous experience with him - then he will undoubtedly ask me - where I am on the priority for CMS! Sorry to take so much time to write this up and send the proposal to you. I just wanted to do my homework of talking to all involved parties in CMS, and gauge their enthusiasm. I have to say that being here (at CERN) during CMS week and talking to relevant people, has at least ensured me that help with the HLT project is needed and is a necessary one to get involved with. I will at least have a chance to really contribute as a newcomer - and that is a nice feeling! I am not sure what the format of these proposals need to be, so you sort of have a rough outline of what I intend to tell the DOE next week. I will use my startup to support the graduate student, and travel, but do need to get DOE to shell out the postdoc funds. As the DOE has told me, by default, the funds I will have will remain at BU and I need to make a strong case to get funding once again!. I do not know all the details of how these things happen, I am learning aeveryday as I make the transition from BostonU to Brown. Thanks for your help. Meenakshi ============================================== bjet tagging: HLT and offline Narain is joining CMS as part of the Brown University group. Brown University group on CMS, under the leadership of Greg Landsberg, is participating in the High Level Trigger (HLT) and Silicon Detector facility via the activities of the FNAL LPC. Narain has agreed to focus her efforts on bjet-tagging, both at the High Level Trigger and the offline event reconstruction software. This emphasizes her expertise in trigger systems (Impact parameter trigger and Level1 track trigger)and bjet .tagging algorithms at the Tevatron and is synergistic with the activities of the Brown group. After consulations with the online selection group and the b/tau PRS group, as the first step, she proposes to implement and study the detail performance of a lepton based bjet-tagger in the High Level Trigger. This algorithm is currently missing from the HLT infrastructure and will be extremely useful as it will be the only tool to assess the performance of the CMS detector for tagging bjets during the first phase of data taking of CMS, when the pixel tracker system is not installed. Eventually this will be merged with the lifetime based tagger in the HLT and be part of the global combined bjet-tagging algorithm. Narain also will be working within the b/tau group to assess the performance of the various btagging algorithms and parameterize the efficiency and mis-tagging rate for bjets as a function of jet kinematics (pT, eta) and detector performance (hit efficiency, number of reconstructed tracks etc) and develop the infrastructure for bjet calibration between data and Monte Carlo. This task will be carried out by implementing the technique .System 8. developed by the Dzero collaboration to characterize the performance bjet algorithms using collider data with very little dependence on Monte Carlos. This project will heavily draw from Narain.s experience on the Tevatron bjet-tagging. The .traditional. method for calibrating bjets at the LHC is based on a sample of ttbar events. This technique provides a complementary method for calibration of bjets during the initially running. It will also be the only method of calibration for extremely high pT jets, above 350 GeV or so (expected in some SUSY and Little Higgs scenarios), as the sample of top quark events become very small for these very high pT region. Narain recently agreed to co-convene the bjet-tagging group based at the FNAL LPC. This group will work with the b/tau PRS group in porting bjet tagging algorithms to CMSSW and also help in understanding the detailed performance of the algorithms in various physics scenarios. We are organizing a workshop at the LPC in near future to introduce users on the details of bjet-tagging in CMSSW, both at the level of .expert developer. and as a .physics user.. Narain requests support for a postdoctoral fellow ($75K) to work on the HLT bjet-algorithm implementation and performance characterization. This will fill a needed gap in the HLT group and help complete the suite of algorithms in HLT, prior to its use by the physics groups in generating a their first trigger menus with preliminary signal efficiencies and expected background rates. -- Prof. Meenakshi Narain e-mail: narain@bu.edu Department of Physics Boston University 590 Commonwealth Avenue phone: (617) 353-6027 Boston, MA 02215 fax: (617) 353-9393