5 Search for Anomalous Higgs Pair Production with CMS

All Life is Problem Solving.

– Karl Popper

In this chapter, the concepts and techniques from the previous sections are applied in the search for non-resonant production of Higgs boson pairs, using data from proton-proton collisions at a centre-of-mass energy of 13 TeV collected in 2016 by the CMS detector at the LHC, corresponding to a total integrated luminosity of \(35.9\ \textrm{fb}^{-1}\). The most probable decay channel for the Higgs boson pairs, where each Higgs boson leads to a \(\textrm{b}\bar{\textrm{b}}\), is considered. While the aforementioned final state is the most frequent by a considerable margin, a large background of similar events is expected from multi-jet QCD processes, which motivates the use of machine learning techniques to construct a summary statistic that can exploit the fine differences between signal and background for statistical inference. In fact, the expected background is so copious that is not possible to generate a sufficiently large number of simulated observations to obtain the required level of modelling accuracy, thus we have to resort to the development of a new data-driven background estimation technique referred to as hemisphere mixing [147]. In addition to setting upper limits on the Standard Model (SM) production of Higgs boson pairs, the data analysis framework is also used to set upper limits in the context of effective field theories (EFT) of anomalous couplings, that parametrise possible deviations from the SM. The main results presented in this section have been carried out within the CMS Collaboration, and have been made public and published [148].