Support vector machine classification on a biased training set: Multi-jet background rejection at hadron colliders

Abstract

This paper describes an innovative way to optimize a multivariate classifier, in particular a Support Vector Machine algorithm, on a problem characterized by a biased training sample. This is possible thanks to the feedback of a signalbackground template fit performed on a validation sample and included both in the optimization process and in the input variable selection. The procedure is applied to a real case of interest at hadron collider experiments: the reduction and the estimate of the multi-jet background in the W → eν plus jets data sample collected by the CDF experiment. The training samples, partially derived from data and partially from simulation, are described in detail together with the input variables exploited for the classification. At present, the reached performance is superior to any other prescription applied to the same final state at hadron collider experiments.

Deep Learning Based Model Identification System Exploits the Modular Structure of a Bio-Inspired Posture Control Model for Humans and Humanoids

“…An a posteriori test can be performed comparing the distribution of the parameters identified on the test set with the expected distribution on real data. This can help the process of choosing between different possible network hyperparameters sets as shown in (Sforza et al, 2011;Sforza and Lippi, 2013). Some parameters were better identified than others as shown in Fig.…”

Measurement of WW and WZ production in the lepton plus heavy flavor jets final state at CDF

“…The analysis strategy consists in selecting a high-acceptance lepton-plus-two-jets sample, using all the selection tools developed for the single-top quark and W H analyses at CDF [2] [4]. A Support Vector Machine (SVM) [5] discriminant is used to suppress the multi-jet (MJ) background. Secondary-vertex jet tagging is applied to enrich the sample in HF and reduce the W + jets background.…”

Measurement of theWWandWZproduction cross section using final states with a charged lepton and heavy-flavor jets in the full CDF Run II data set

“…However, this enhances the background contribution from QCD multijet events classified as having a W -boson-like signature, e.g., if a particle in a jet meets the lepton identification criteria and moderate E T is generated from energy mismeasurement. To cope with this, a multivariate multijet rejection strategy based on a support-vector-machine algorithm is developed [22].…”

“…The specific choices used in this analysis are described in more detail in Ref. [22]. The input training sets used are simulated W → eν+jets signal events and MJbackground events from data, as described in Sec.…”

“…Both requirements strongly suppress the background of ZZ production where one of the two bosons decays hadronically and the other leptonically. A support-vector-machine (SVM) algorithm [22] is used to select events consistent with W → ν+jets production. Finally, the hadronic decay of the W or Z boson is identified by requiring events with two jets with a large component of the momentum transverse to the beam (transverse momentum) and where, in at least one of them, a secondary-decay vertex is identified in at least one jet, indicating the presence of a b or c hadron (HF tag).…”