Search for supersymmetry in events with opposite-sign dileptons and missing transverse energy using an artificial neural network

In this paper, a search for supersymmetry (SUSY) is presented in events with two opposite-sign isolated leptons in the final state, accompanied by hadronic jets and missing transverse energy. An artificial neural network is employed to discriminate possible SUSY signals from a standard model background. The analysis uses a data sample collected with the CMS detector during the 2011 LHC run, corresponding to an integrated luminosity of $4.98{\mathrm{fb}}^{-1}$ of proton-proton collisions at the center-of-mass energy of 7 TeV. Compared to other CMS analyses, this one uses relaxed criteria on missing transverse energy (${\mathrm{E̸}}_{\mathrm{T}}>40\mathrm{GeV}$) and total hadronic transverse energy ($H_{\mathrm{T}}>120\mathrm{GeV}$), thus probing different regions of parameter space. Agreement is found between standard model expectation and observations, yielding limits in the context of the constrained minimal supersymmetric standard model and on a set of simplified models.

Figure 1

Extrapolation factor $R_{\mathrm{ext}}$, as obtained from simulated events.

Figure 2

Data (black points with error bars) vs simulated events (red bands) comparison of the ANN output distributions in the various control regions. Top: ANN output in the control region used to perform the extrapolation with systematic uncertainties included (left), and the ratio between data and simulated events (right) with both statistical (black error bars) and systematic (red bands) uncertainties shown. Bottom: main control region subdivided into the following two to perform cross-checks: $t\overline{t}$ enriched control region A (left), $Z+\mathrm{\text{jet}}$ enriched control region B (right) with systematic uncertainties included.

Figure 3

Left: ANN output for the SM background (red bands) and SMS low ${\mathrm{E̸}}_{\mathrm{T}}$-low $H_{\mathrm{T}}$ events (blue line). Right: efficiency (red) and purity (blue) vs the minimum ANN output value for SMS low ${\mathrm{E̸}}_{\mathrm{T}}$-low $H_{\mathrm{T}}$ events in the signal region.

Figure 4

Distributions in seven input ANN variables for simulated SM background events (red shaded band showing combined statistical and systematic uncertainties) and SMS low-${\mathrm{E̸}}_{\mathrm{T}}$ low-$H_{\mathrm{T}}$ events (blue histogram), normalized to the same number of events. The data are also shown for comparison (black points with error bars).

Figure 5

ANN output for the data (black points) and the SM prediction based on control regions in data (gray band) in the signal region. The uncertainty on the SM prediction includes both statistical and systematic components.

Figure 6

Distribution in ${\mathrm{E̸}}_{\mathrm{T}}$ (top left) and $H_{\mathrm{T}}$ (top right) for signal-like events (ANN $\mathrm{\text{output}}>0.95$) for data (black points with error bars) and simulation (red shaded band). The ${\mathrm{E̸}}_{\mathrm{T}}$ vs $H_{\mathrm{T}}$ distributions for signal-like events ($\mathrm{ANN}>0.95$) in data (bottom left) and simulation (bottom right); the regions to the right and up of the black line are the ones also explored by other CMS opposite-sign dilepton searches.

Figure 7

95% C.L. exclusion limits on the simplified model scenarios with the ANN analysis. The acceptance (fraction of events surviving event selection and candidate event selection) $\times$ efficiency (fraction of events surviving ANN selection) (top) and 95% C.L. upper cross section limit (bottom) are shown for different gluino and neutralino masses. The region just below the diagonal is not considered due to inadequate initial state radiation modeling.

Figure 8

Expected (blue) and observed (red) 95% C.L. exclusion limit for the ANN analysis (for ANN $\mathrm{\text{output}}>0.95$) in the CMSSM plane. The one $\sigma$ experimental error around the expected limit and the one $\sigma$ theoretical error around the observed limit are also shown.