Search for the Type III Seesaw Mechanism with Multivariate Analysis

The vanishingly small mass of neutrinos continues to be an open problem in high energy physics. The Type III Seesaw Mechanism is a model which predicts a triplet of heavy fermions with high masses that offset the low masses of neutrinos. We search for evidence of these heavy Seesaw fermions in the data taken at the CMS detector of the Large Hadron Collider (LHC), restricting our search to only those events which result in 3 or more leptons in the final state. Past methods involved binning data in kinematic variables of the collisions and comparing to simulations of both standard model backgrounds and signal of a new model. Previous limits calculated using one-dimensional binning schemes excluded Seesaw up to a mass of 840 GeV with an observed cross section limit of .05 pb but required an observed cross section limit of .2 pb to exclude the 200 GeV masspoint as the sensitivity of of one-dimensional binning schemes declines for low masses. This study considers the application of Boosted Decision Trees (BDTs) as a method to improve sensitivity in the background dominated low mass regime. For data binned in the new BDT variable, the observed cross section limit for the 140 GeV masspoint is reduced by 57%.