Inaugural ATLAS Searches for Resonant Di-Higgs and SH signals in the Boosted, Fully-hadronic bbVV Final State at ATLAS using √s = 13 Tev data and Novel Machine Learning Techniques.

In this work two distinct yet intimately related efforts are presented. First, the development of a jet tagger consisting of a hybrid, parameterized convolutional neural network optimized for discriminating boosted, four prong jets against a majority QCD background. Uncertainties in tagging four prong jet are estimated, calibrating the tagger for use on such objects in data for the first time in ATLAS. Second, the aforementioned tagger is used as a core component in an analysis to set 95% CL limits for resonant di-Higgs production into the fully hadronic bbV V final state. This analysis is reinterpreted under the X →SH model into the same final state.