1 / 23

Statistical issues in Particle Physics analyses

Statistical issues in Particle Physics analyses. Louis Lyons Imperial College and Oxford CMS and CDF experiments l.lyons@physics.ox.ac.uk Ovronnaz, Sept 2009. Lecture topics. Introduction to Particle Physics Learning to love the Covariance Matrix 2) Do’s and Dont’s with Likelihoods

ide
Télécharger la présentation

Statistical issues in Particle Physics analyses

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Statistical issues in Particle Physics analyses Louis Lyons Imperial College and Oxford CMS and CDF experiments l.lyons@physics.ox.ac.uk Ovronnaz, Sept 2009

  2. Lecture topics • Introduction to Particle Physics Learning to love the Covariance Matrix 2) Do’s and Dont’s with Likelihoods 3) Statistical issues concerning searches for New Physics: exclusions and discoveries

  3. Particle Physics • What it is • Typical experiments • Typical data • Typical analysis

  4. What is it? Search for ultimate constituents of matter Ancient tradition What constitutes acceptable theory? “Number of fundamental entities at most A F E W ” Early ideas due to Greeks

  5. What is it? Search for ultimate constituents of matter Ancient tradition What constitutes acceptable theory? “Number of fundamental entities at most A ir F ire E arth W ater ”

  6. How many? Greeks’ AFEW 4 Dalton’s atoms ~20  120 e and p 2 and n, π, μ, Ω… >1000 “Elementary” particles Quark model 3  35 Future substructure? 1?

  7. How big? Human ~ 1 metre Dust ~10-5 metres Atom ~ 10-10 metres Nucleus ~ 10-15 metres Quarks < 10-18 metres Mass of proton = 1 GeV neutrino < 2 eV top = 175 GeV

  8. Forces of nature • Gravity (mediated by Gravitons, ) • Electromagnetism (Photons, γ ) • Nuclear strong force (Gluons, g) • Nuclear weak force (Int Vect Bosons, W Z)

  9. Elementary particles: Force carriers: photon, W and Z, graviton, gluons Higgs for the masses LEPTONS: e μτ and neutrinos Composite particles: HADRONS: Made from qqq or q q e.g. proton = uud, π+ = u d

  10. Typical Experiments • Experiment Energy Beams # events Result • LEP 200 GeV e+ e- 107 Z N = 2.987± 0.008 • BaBar/Belle 10 GeV e+ e- 108 B anti-B CP-violation • Tevatron 2000 GeV p anti-p “1014” SUSY? • LHC 14000 GeV p p (2010…) Higgs? • KK ~3 GeV μ 100 ν oscillations

  11. In the LHC tunnel

  12. ATLAS Detector at LHC 45 metres long, 25 m high. 3000 physicists from 40 countries

  13. CDF at Fermilab

  14. Simulated SUSY event in ATLAS Jet Muons Jet Missing ET

  15. Interesting Physics is rare

  16. TypicalAnalysis • Parameter determination: dn/dt = 1/τ * exp(-t / τ) Worry about backgrounds, t resolution, t-dependent efficiency • 1) Reconstruct tracks • 2) Select real events • 3) Select wanted events • 4) Extract t from L and v • 5) Model signal and background • 6) Likelihood fit for lifetime and statistical error • 7) Estimate systematic error τ± στ(stat)± στ(syst) Goodness of fit

  17. Separate signal from background Typical x: masses energies missing E b-tag event shape isolated γ ……. p(data|H0) x2 p(data|H1) x1 Use cuts, decision trees, Fisher discriminants, NN, SVM, boosting, bagging,….. All aim to approximate optimal likelihood ratio ordering

  18. Width of Z0 3 light neutrinos

  19. TypicalAnalysis Hypothesis testing: Peak or statistical fluctuation? • Why 5σ ? • Past experience • Look elsewhere • Bayes priors

  20. Read more LL, “Open statistical issues in Particle Physics”, Annals of Applied Statistics 2 (2008) 887 PHYSTAT-LHC and earlier workshops CDF Statistics Committee web-page

More Related