中央研究院 Logo

In this talk we discuss a recent search over the parameter space of the Next-to-Minimal Supersymmetric Standard Model using deep learning based techniques. The first part focuses on using deep learning to fit scalars with
masses of 95 GeV and 650 GeV, motivated by recent experimental results, as well as discrepancies in Electro-Weakino searches. In the second part we demonstrate the phenomenological relevance of this search. We use a proposed neural network architecture to improve selection of events in future dark matter searches with mono-H and mono-Z signatures at the HL-LHC, with promising results.