Learning to increase matching efficiency in identifying additional b-jets in the $$\text {t}\bar{\text {t}}\text {b}\bar{\text {b}}$$ process

Abstract

AbstractThe$$\text {t}\bar{\text {t}}\text {H}(\text {b}\bar{\text {b}})$$tt¯H(bb¯)process is an essential channel in revealing the Higgs boson properties; however, its final state has an irreducible background from the$$\text {t}\bar{\text {t}}\text {b}\bar{\text {b}}$$tt¯bb¯process, which produces a top quark pair in association with a b quark pair. Therefore, understanding the$$\text {t}\bar{\text {t}}\text {b}\bar{\text {b}}$$tt¯bb¯process is crucial for improving the sensitivity of a search for the$$\text {t}\bar{\text {t}}\text {H}(\text {b}\bar{\text {b}})$$tt¯H(bb¯)process. To this end, when measuring the differential cross section of the$$\text {t}\bar{\text {t}}\text {b}\bar{\text {b}}$$tt¯bb¯process, we need to distinguish the b-jets originating from top quark decays and additional b-jets originating from gluon splitting. In this paper, we train deep neural networks that identify the additional b-jets in the$${\text {t}}{\bar{\text {t}}}{\text {b}}{\bar{\text {b}}}$$tt¯bb¯events under the supervision of a simulated$$\text{t}\bar{\text{t}}\text{b}\bar{\text{b}}$$tt¯bb¯event data set in which true additional b-jets are indicated. By exploiting the special structure of the$$\text {t}\bar{\text {t}}\text {b}\bar{\text {b}}$$tt¯bb¯event data, several loss functions are proposed and minimized to directly increase matching efficiency, i.e., the accuracy of identifying additional b-jets. We show that, via a proof-of-concept experiment using synthetic data, our method can be more advantageous for improving matching efficiency than the deep learning-based binary classification approach presented in [1]. Based on simulated$$\text {t}\bar{\text {t}}\text {b}\bar{\text {b}}$$tt¯bb¯event data in the lepton+jets channel from pp collision at$$\sqrt{s}$$s= 13 TeV, we then verify that our method can identify additional b-jets more accurately: compared with the approach in [1], the matching efficiency improves from 62.1$$\%$$%to 64.5$$\%$$%and from 59.9$$\%$$%to 61.7$$\%$$%for the leading order and the next-to-leading order simulations, respectively.