Search for Standard Model Higgs Boson Production in Association with a W Boson using Matrix Element and Boosted Decision Tree Techniques with 2.7 fb-1 of CDF Data

Barbara Alvarez2, Florencia Canelli1, Bruno Casal Laraña2, Javier Cuevas Maestro2, Peter Dong3,
Craig Group1, Enrique Palencia1, Alberto Ruiz2, Bernd Stelzer3, Rocio Vilar2, Rainer Wallny3
 
1Fermilab, 2IFCA (CSIC-UC), 3UCLA

 


- Abstract -
- Event Selection -
- Some Input Variables -
- Cross-Checks -
- Systematics -
- Results -
- Public Note -




Abstract:

We present a search for Standard Model Higgs boson production in association with a W boson using 2.7 fb-1 of CDF II data collected between 2002 and 2008. This search is performed using two complementary techniques: a matrix element technique is used to calculate event probability densities for the signal and background hypothesis and a multivariate technique, based on Boosted Decision Tree, is used to combine the event probability densities with kinematic variables to build a final discriminant distribution which is fitted to the data using a binned likelihood approach. We observe no evidence for a Higgs boson signal and set 95 % confidence level upper limits on the WH production cross section times the branching ratio of the Higgs boson to decay to bb-bar pairs of &sigma(pp-bar -> WH)xBR(H -> bb-bar)/SM < 3.52 to 107.6 for Higgs boson masses between mH=100 GeV/c2 and mH=150 GeV/c2 . The expected (median) limit estimated in pseudo-experiments is: &sigma(pp-bar -> WH)xBR(H -> bb-bar)/SM < 4.07 to 52.1 at 95 % C.L.




Event Selection:

This analysis uses events from leptonic decay of the W boson. We require a single, well isolated high-transverse-energy lepton, large missing transverse energy (from the neutrino), and exactly two high-transverse-energy jets. Of these jets, we require at least one to be identified as originating from a b-quark by secondary vertex tagging. The secondary vertex tag identifies tracks associated with the jet originating from a vertex displaced from the primary vertex. We further require the missing transverse energy and the jets not to be collinear for low values of missing transverse energy. This requirement removes a large fraction of the non-W background while retaining most of the signal.
Our major backgrounds come from W + heavy flavor jets, Wbb-bar, Wcc-bar, and Wc+jet; mistags which are W + light quark/gluon events that are mistakenly tagged as b-jets due to detector resolution effects; Non-W, which are mostly multijet events in which a jet is mistakenly identified as a lepton and jets are mismeasured, providing a false missing transverse energy signature; and top pair production events in which one lepton or two jets are lost due to detector acceptance.

Number of expected signal and background events, in the 2 jet bin, in 2.7 fb -1 of CDF data, passing all event selection requirements





Some Input Variables:

Some of the variables used for the training of the BDT in the 2 jet bin channels are listed and ploted (left: W+2jets zero tags, midle: W+2jets 1 tag, right: W+2jets 2 tags) below:
-- HT --


-- ET(j1) --


-- ET(j2) --


-- Mj1j2 --


-- EPD1tag --


-- EPD2tag --


-- log(PWH/Pschan) --


-- log(PWH/PWbb) --


-- log(PWH/Ptt) --







Cross-Checks:

In addition to validate the input variables of the BDT, we evaluate the BDT outputs in the untagged sample. In all control samples, the data agrees well with the Monte Carlo prediction.

Distribution of the BDT output trained for W+2jets-1tag evaluated in the untagged sample.

Distribution of the BDT output trained for W+2jets-2tag evaluated in the untagged sample.






Systematics:

We address systematic uncertainty from several different sources:
  • jet energy scale
  • initial state radiation
  • final state radiation
  • parton distribution functions
  • luminosity
  • b-tagging SF
Systematic uncertainties can influence both the expected event yield (normalization) and the shape of the discriminant distribution. Normalization uncertainties are estimated by recalculating the acceptance using Monte Carlo samples altered due to a specific systematic effect. The WH normalization uncertainty is the difference between the systematically shifted acceptance and the default one and are shown in the next Table.


Systematic uncertainty Single Tag Double Tag
Jet energy scale 2.0 % 2.0 %
ISR/FSR + PDF 3.1 % 5.6 %
Lepton ID ~2.0 % ~2.0 %
Luminosity 6.0 % 6.0 %
b-tagging SF 3.5 % 8.4 %

Rate Systematic Uncertainties in single and double tagged samples.




Results:

Distributions of the BDT output trained for single (left) and double (right) tag evaluated in the single (left) and double (right) tagged sample.



In order to extract the most probable WH signal content in the data we perform the maximum likelihood method. We have analyzed 2.7 fb-1 of CDF Run II data. We observe no evidence for a Higgs boson signal and set 95 % confidence level upper limits on the WH production cross section times the branching ratio, in SM units, of the Higgs boson to decay to bb-bar pairs of &sigma(pp-bar -> WH)xBR(H -> bb-bar)/SM < 3.52 to 107.6 for Higgs boson masses between mH = 100 GeV/c2 and mH = 150 GeV/c2. The expected (median) sensitivity estimated in pseudo experiments is:
&sigma(pp-bar -> WH)xBR(H -> bb-bar)/SM < 4.07 to 52.1 at 95 % C.L.




mH (GeV) Expected (&sigma/SM) Observed (&sigma/SM)
100 4.07 3.52
105 4.28 4.19
110 4.82 6.67
115 5.64 5.75
120 7.02 7.98
125 8.32 8.36
130 10.7 11.7
135 14.2 20.5
140 20.8 26.1
145 29.4 54.5
150 52.1 107.6

Expected and observed upper limit cross sections in SM units for different Higgs mass points


mH (GeV) Expected (&sigma (pb)) Observed (&sigma (pb))
100 0.944 0.817
105 0.860 0.842
110 0.815 1.127
115 0.767 0.782
120 0.730 0.830
125 0.691 0.694
130 0.674 0.737
135 0.639 0.923
140 0.624 0.783
145 0.588 1.090
150 0.625 1.291

Expected and observed upper limit cross sections for different Higgs mass points




Public Note:

For more details about the analysis you can have a look at our public note




-- email us