New Physics at the LHC: Direct and Indirect Probes

This thesis presents the results for two searches for new physics performed with the ATLAS experiment. The first, a search for the rare B-meson decay $B_s \rightarrow \mu \mu$ and measurement of its branching ratio, uses 25 fb$^{−1}$ of $\sqrt{s} =$ 7 and 8 TeV data recorded during 2011 and 2012. After observing a small number of these decays, a branching ratio of $\mathcal{B}(B_s \rightarrow \mu \mu) = (0.9^{+1.1}_{-0.8}) \times 10^{−9}$ is measured, assuming non-negative event yields. This is compatible with the Standard Model at the $2 \sigma$ level. The second, a search for direct pair production of the supersymmetric top quark partner, is performed using 36.07 fb$^{−1}$ of $\sqrt{s} =$ 13 TeV data recorded during 2015 and 2016. Final states with a high jet multiplicity, no leptons and large missing transverse momentum are selected to target these decays, with several signal regions designed to cover a wide range of particle masses. No excess is observed, with all signal regions being compatible with the Standard Model within $2 \sigma$. Limits are set on the stop mass, excluding up to $m_{\bar{i}_1} =$ 940 GeV for values of $m_{\bar{\chi}^0_1}$ below 160 GeV, assuming a 100% branching fraction to $\bar{t}_1 \rightarrow t \bar{\chi}^0_1$ decays. In addition two reinterpretations of this data are presented, for a gluino-mediated stop production scenario and a direct dark matter production scenario. No excess is observed for either model, and limits are set on the mass of the relevant particles. Finally a viability study into using machine learning techniques to improve on existing SUSY search methods has been performed, with the initial results proving promising.