Differences

This shows you the differences between two versions of the page.

--- 2015:groups:sm:qg [2015/06/03 15:11]
jesse.thaler created
+++ 2015:groups:sm:qg [2015/06/06 08:36]
simon.pltzer [Quark/Gluon Enrichment Studies]
@@ Line 1: / Line 1: @@
+====== Quark/Gluon Enrichment Studies ======
+a.k.a. Hunting the White Whale of Jet Substructure
+  * Andy Buckley
+  * Jon Butterworth
+  * Mario Campanelli
+  * Marat Freytsis
+  * Peter Loch <loch@physics.arizona.edu>
+  * Deepak Kar
+  * Simon Plätzer
+  * Andrzej Siodmok <andrzej@cern.ch>
+  * Peter Skands <peter.skands@monash.edu>
+  * Dave Soper
+  * Gregory Soyez
+  * Frank Tackmann
+  * Jesse Thaler <jthaler@mit.edu>
+  * ...
-====== Original Notes from Gregory ======
+Link to GitHub repository:  https://github.com/gsoyez/lh2015-qg
+===== Preliminaries =====
+==== Quark/gluon discrimination well-defined? ====
+  * Of course, at the hadron level, you can't define a quark jet vs. a gluon jet unambiguously.
+  * That said, one can talk about quark/gluon enriched samples, where restrictions are placed on the final state to preferentially select quark- or gluon-initiated jets (e.g. gluon enrichment in dijets, quark enrichment in vector boson plus jet).
+  * In fixed-order QCD, there is an ambiguity from soft gluon splitting to wide-angle quark/anti-quark.  However, in the eikonal limit, there is no ambiguity (up to power corrections), so quark/gluon calculations can be done at the parton level in the eikonal limit (relevant for resummed calculations).
+  * If needed, we can use flavored jet algorithms to give an IRC safe definition of jet flavor at the parton level.
+==== How to isolate quark vs. gluon samples? ====
+  * Ultimately, we need an operational definition of quark- and gluon-enriched samples (e.g. event type, rapidity correlations, event shapes).
+  * This will allow us to separate the **measurement** of jet properties from the **interpretation** of those properties in the context of discrimination/enrichment studies.
+  * One has to be aware of process dependence, since a quark in one context may not look like a quark in another context (color correlations).
+  * Ultimately, need MC studies to compare to behavior in data.
+==== "Discrimination" really the right word? ====
+  * Probably better to talk about "quark/gluon enrichment".
+  * For physics applications, we want to achieve S/sqrt{B} improvement, which isn't really the same as quark/gluon discrimination.
+  * Similar issues arise in how to define a "hadronic W".
+  * Quark/gluon enrichment should be a piece of a more refined analysis.
+  * We can provide general recipes, but should not aim for optimal analyses, which are only sensible in the context of specific physics goals.
+==== What is the killer app of quark/gluon enrichment? ====
+  * VBF tagging
+  * Rejecting (stochastic) pileup jets (important for VBF)
+  * SUSY multi-jet tends to be quark-enriched
+  * Enhancing W/Z/t/H in moderately boosted regime (where we can quark-tag the subjets.
+===== Physics Issues =====
+==== Separating final state from initial state effects ====
+  * Different jet shapes probe different phase space regions.  For example, jet mass is more sensitive to wide angle physics while multiplicity is more sensitive to collinear physics.
+  * Differences between MC programs appear in multiplicity-like observables, so most likely a final state effect.
+  * We can probe different physics by looking at hard core (collinear, FSR, beta -> 0) vs. wide angle (soft, ISR, beta -> infty).
+==== FSR effects ====
+  * FSR effects should be dominant at small angles, yielding universal properties.
+  * Tuning of gluon final state shower can affect jet shapes.
+  * Examples:  g -> q qbar vs. g -> gg, including spin-polarization information
+  * Do beyond-LL effects help or hurt quark/gluon enrichment?
+  * What about the impact of heavy flavor?
+==== ISR effects ====
+  * ISR effects should dominate at large angles
+  * Highly process dependent, depends on color corrections of jet with ISR
+  * We will attempt to deemphasize these in our study, if possible
+==== Experimental Results ====
+  * ATLAS paper suggests that beta -> 0 (i.e. hard core) is not as effective as NLL calculations suggest.  (see http://arxiv.org/abs/1405.6583 Appendix A.)
+  * CMS finds ptD (an example of a beta -> 0 observable) is quite effective.
+  * ATLAS sees considerable process dependence, whereas CMS has not emphasized this issue.  Is this connected to ISR in some way?
+  * ATLAS A14 tune already uses jet shapes, and finds that alpha_s has to be tuned downward in Pythia 8.  This, however, has a detrimental effect on LEP measurements, so one has to be cautious about this.
+  * Is there a tuning flat direction?
+==== Hemisphere quark/gluon definitions in e+e- ====
+  * Consider the case of e+ e- -> q qbar.  Partition event into (thrust) hemisphere, define hemisphere flavor by summing over flavors of hemisphere constituents.
+  * At LO, we can unambiguously define hemisphere flavors.
+  * At NLO, we can also unambiguously define flavor via hemisphere, though there is now a small gluon fraction from gluon recoiling against q qbar pair.
+  * At NNLO, things are more complicated.
+    * Can have soft gluon splitting into q-qbar in different hemispheres, creates IRC safety issue.
+    * One can use a flavored algorithm (BSZ) to define the flavour of two flavor-kt jets
+  * Ultimately, want to give an operational definition of flavor based on the Born-level operator contributing to the process.
+    * Claim:  all subtleties are formally power suppressed.
+    * Use case, VBF, two jets with a third jet veto, q/g well-defined in the exclusive limit.
+==== Flavored Jet Algorithms ====
+  * This is a topic worthy of its own Les Houches study.
+  * For pp collisions, multiple possible uses of flavored jet algorithms.
+  * One can just run flavor-kT
+  * Or one can run flavor-kT to define flavor ghosts, and run standard anti-kT.
+  * Or one can run flavor-kT for deflavoring constituents, and then run standard anti-kT.
+===== Ultimate Goal for Les Houches Study =====
+  * Recommendation to ATLAS and CMS for observables that should be measured which carry quark/gluon information.
+  * These observables must be defined on the final state alone (i.e. fiducial cross section).
+  * These observables should help enrich quarks over gluons (or vice versa).
+  * Eventually, these observables should be useful for MC tuning, with controllable systematics.
+  * Make recommendation about robustness vs. performance.  We will likely emphasize robustness, since performance depends strongly on process dependence, pileup.
+  * Question:  How should we discuss low pT vs. high pT
+===== Initial Les Houches Study =====
+==== Key Question ====
+  * Do we understand FSR modeling by workhorse parton showers?
+  * Start with the clean case of e+e-, move to pp later.
+==== Basic Plan ===
+  * Take e+e- -> q qbar, and e+e- -> g g
+  * Vary collision energy, jet radius
+  * Choose a core set of jet shapes
+  * Use as many MC options as possible.
+  * Question: use ROC curves or mutual information (I(T;A)) to quantify discrimination power?
+    * Answer: doesn't really matter, probably I(T;A) is easier to begin with.
+    * Better answer:  Use separation (S-B)^2 / (2 (S + B)).
+==== Core Jet Shapes ====
+  * Generalized angularities (kappa, beta)
+    * (1,0.5)
+    * (1, 1) -- jet width
+    * (1,2)  -- jet mass
+    * (0,0) -- multiplicity
+    * (2,0) -- ptD
+  * Question:  Apply on full events or just tracks
+    * Answer:  Apply on full events.
+  * Question:  Choice of axes? (issue of recoil)
+    * Answer:  WTA recombination axes from anti-kT
+  * Question:  Sum over particles (angularity-style) vs. sum over pairs (ECF-style)
+    * Answer:  Sum over particles (angularity-style)
+  * Question:  Plot linear or log scale?
+    * Answer:  Do both if it makes sense, better for angularities to have log scale.
+==== Supplemental Jet Shapes ====
+  * More generalized angularities (kappa, beta)
+    * (0.5,0.5)
+    * (0.5,1.5) -- should give bad performance
+    * limit (1+epsilon,0) / epsilon -- should give good perfomance
+  * Ellipticity/eccentricity
+  * Covariance matrix observables
+  * Pull
+  * Psi(r) -- the jet shape
+  * Check Gallicchio and Schwarz catalog
+  * tau21, or ECF(2,3)
+  * Generalized angularities with soft-drop jets, varying beta_SD
+  * Do sum over pairs version of angularities (i.e. ECF-style)
+==== Analysis Workflow ====
+  * Rivet analysis in place which computes from a HepMC event sample the various generalised angularity distributions.
+  * Processes to consider:
+    * mu+mu- -> spin1 -> q qbar  take photons
+    * mu+mu- -> spin0 -> g g  take Higgs
+    * for tests of universality:  mu+mu- -> spin0 -> q qbar
+  * Energies
+    * Q=sqrt{s} = 50, 200, 800 GeV
+    * Optionally:  Q = 100, 400 GeV
+  * Jet definition:
+    * ee-antikt [genkt, p=-1], WTA_modp recomb scheme
+    * R = 0.3, 0.6, 0.9
+  * Add thrust from thrust hemispheres for anticipated analytic comparisons
+  * Add multiplicity (event-wide) in bins of thrust:
+     * T < 5 GeV/sqrt(S)
+     * 5 GeV/sqrt(S) < T < 0.1
+     * 0.1 < T < 0.2
+     * 0.2 < T
+==== Preliminary plots for meeting on Thursday ====
+{{:2015:groups:sm:ga_10_20.pdf|}}
+{{:2015:groups:sm:ga_10_10.pdf|}}
+{{:2015:groups:sm:ga_10_05.pdf|}}
+{{:2015:groups:sm:ga_00_00.pdf|}}
+{{:2015:groups:sm:ga_20_00.pdf|}}
+==== Questions ====
+  * Is discrimination power (e.g. for width) coming from the hadronization regime?
+    * Possibility:  Isolate hadronization regime (thrust ~ LambdaQCD/Q) and shower regime (thrust ~ 0.1-0.2) and optionally hard jet regime (thrust >~ 0.25).  Study scaling of, e.g., multiplicity as a function of Q in each of these regimes.
+  * By testing pythia vs. herwig, can we test string vs. cluster hadronization?
+  * Is there jet radius dependence?
+  * Does matching help in controlling quark/gluon uncertainties?
+  * Universality/process dependence of conclusions?
+    * Related to whether the discrimination power comes from the core or the periphery of jet.
+===== Next Les Houches Study (for after LH) =====
+  * Above study at hadron colliders, using dijets, W/Z/gamma + j, and maybe t tbar samples
+===== Analytic Les Houches Study? =====
+  * Analytic predictions known/available/straightforward for:
+    * Quark thrust: N^3LL' + N^3L0
+    * Gluon thrust: N^2LL' + N^2L0
+    * ang (kappa =1):  NLL'
+  * Can we do useful quark/gluon study from analytic results?
+===== Notes from Tuesday Meeting =====
 <code>
@@ Line 134: / Line 342: @@
 </code>
+===== Notes from Thursday Meeting =====
+<code>
+Meeting in Les-Houches
+Presentation of the wiki notes: list of contributors, ...
+Presentation of the status of the software:
+  start w e+e- and do pp later
+  Rivet analysis in place which computes from a HepMC event sample the various generalised angularity distributions
+Reminder: what we mean by a q and a g is e+e-\to qq and e+e-\to gg
+  If we want to do something more refined:
+    - at LO we can unambiguously sum flavours in hemispheres defined by thrust
+    - at NLO we can unambiguously sum flavours in hemispheres defined by thrust
+      we get a quark and a small gluon fraction
+    - at NNLO things are more complicated. We can use a flavoured
+      algorithm (BSZ) to define the flavour of each hemisphere
+    - for pp collisions, we should use a flavoured algorithm to
+      determine flavour, and then find a way (e.g. using ghosts) to
+      run anti-kt jets. This would deserve a topic per se (a LH accord)?
+    - Question: can we match to the Born and find an operatiroal
+      definition up t power corrections?
+      Use case: VBF, two jets with a third jet veto. q/g well-defined
+                in the exclusive limit
+Questions to look into:
+ - is the discrimination power (e.g. for width) coming from the hadronisation regime?
+ - plotting in log binninb?
+ - pythia v. herwig important to test string v. cluster hadronisation
+ - isolate hadronisation regime. Study the scaling in different bins
+   of one angularity (e.g. thrust). Take a hadronisation region
+   (T\propto LQCD/Q) and a shower region (T~0.1-0.2) plus optionally a
+   "hard jet region" (T >~ 0.25)
+ - does mathing help?
+ - jet radius dependence (edit analysis and recompile)
+ - analytic predictions?
+    for thrust: ee->qq known at N^3LL' + N^3LO
+                ee->qq known at N^2LL' + N^2LO
+        ang(bkappa=1): NLL'
+ - question of the universality/process dependence of the conclusions?
+   Related to whether the power comes from the core or the periphery?
+ - process to consider:
+     mu+mu- -> spin1 -> qq  take photons
+     mu+mu- -> spin0 -> gg  take Higgs
+   for tests of universality
+     mu+mu- -> spin0 -> qq
+ - Energies Q=sqrt = 50, 200, 800 GeV
+   jetdef: ee-antikt [genkt, p=-1], WTA_modp recomb scheme
+   radii: 0.3, 0.6, 0.9
+ - add thrust from thrust hemispheres for analytic purpose
+ - add multiplicity (event-wide) in bins of thrust:
+     T < 5 GeV/sqrt(S)
+GeV/sqrt(S) < T < 0.1
+.1 < T < 0.2
+.2 < T
+</code>
+===== Notes for Jesse for Preparing Summary Talk =====
+  * Quark is more of an adjective than a noun.
+  * Pseudo-quark?  (That language doesn't go over very well.)
+==== What is a Quark Jet? ====
+(From ill-defined to well-defined)
+  * A quark parton
+  * A Born-level quark parton
+  * The initiating quark parton in a final state shower
+  * An eikonal line with baryon number 1/3 and carrying triplet color charge
+  * A quark operator that appears in a hard matrix element in the context of a factorization theorem.
+  * A parton-level jet object that has been tagged as a quark using a soft-safe flavored jet algorithm (automatically collinear safe if you sum constituent flavors).
+  * A phase space region (as defined by an unambiguous hadronic fiducial cross section measurement) that yields an enriched sample of quarks (as interpreted by some suitable, though fundamentally ambiguous, criterion).
+(Sometimes people think we care about the top of the list while we are really focused entirely on the bottom.)

Les Houches

User Tools

Site Tools

Differences

Page Tools