1 / 15

Proposal to Upgrade the MIPP Experiment-P960

Proposal to Upgrade the MIPP Experiment-P960. Rajendran Raja Nov 4-2010 Format of talk. NuMI Target analysis. Event Displays. Acceptances and momentum resolutions. MIPP detector responses. Need to know particle content a priori to calculate likelihoods. Bayes ’ Theorem based algorith m.

gilda
Download Presentation

Proposal to Upgrade the MIPP Experiment-P960

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. Proposal to Upgrade the MIPP Experiment-P960 Rajendran Raja Nov 4-2010 • Format of talk Rajendran Raja, Fermilab PAC

  2. NuMI Target analysis Rajendran Raja, Fermilab PAC

  3. Event Displays Rajendran Raja, Fermilab PAC

  4. Acceptances and momentum resolutions Rajendran Raja, Fermilab PAC

  5. MIPP detector responses Rajendran Raja, Fermilab PAC

  6. Need to know particle content a priori to calculate likelihoods • Bayes’ Theorem based algorithm Rajendran Raja, Fermilab PAC

  7. Bayes' theorem – Global PID formalism • The joint probability P(H,x) can be written as (H = e,p,K,p; x = dE/dx, ToF, rRICH,...)where P(H) is the probability of a particular hypothesis. This is what we are trying to determine. These equations are for a given momentum. We have suppressed the momentum dependence for simplicity. • By Bayes' theorem • This leads to • We determine P(H) iteratively. Assume that all hypotheses are equally likely initially, i.e. P(H) = ¼ since there are 4 hypotheses (e/p/K/p). For each track, we then determine the posterior probability P(H|x) which is used to weight the track for each hypothesis. • The resulting P(H) is used for the next iteration, till convergence. • The aim is not to determine whether each particle is definitely one type or the other but to determine the maximum likelihood momentum functions for each hypothesis. Each particle enters all hypotheses plots with its appropriate hypothesis dependent weight. • We treat MC and data as two separate experiments, each with slightly different behavior. We test the algorithm on the MC, since we know the answer. – (Movie) Rajendran Raja, Fermilab PAC

  8. Green GlobalPid. Red MCTRUTH 15 iterations Positive charges Rajendran Raja, Fermilab PAC

  9. Green GlobalPid. Red MCTRUTH 15 iterations Negative charges Rajendran Raja, Fermilab PAC

  10. Data p vs pt positives Rajendran Raja, Fermilab PAC

  11. Data p vs pt negatives Rajendran Raja, Fermilab PAC

  12. Comparison of Data and MC for negatives Rajendran Raja, Fermilab PAC

  13. Comparison of Data and MC for Positives Rajendran Raja, Fermilab PAC

  14. Comparison of Monte Carlo Positive and negative spectra Rajendran Raja, Fermilab PAC

  15. Comparison of Data Positive and negative spectra Rajendran Raja, Fermilab PAC

More Related