Research Projects

Supersymmetric theories are one of the most promising and well-studied (theoretically, phenomenologically as well as experimentally) beyond the Standard Model (BSM) scenarios because of it’s ability to address several theoretical and observational issues of the SM. The particle spectrum of R-parity conserving (RPC) supersymmetric scenarios includes a stable, weakly interacting massive particle which can be a viable candidate for cold dark matter and explain the observed dark matter relic density. The presence of supersymmetry at the TeV scale can explain several donating theoretical issues like the hierarchy/naturalness problem, unification of the gauge couplings etc. As a result, one of the prime objectives of the large hadron collider (LHC) experiment is to search for sparticles. Experimental groups in both CMS and ATLAS Collaboration are dedicated to searching for sparticles. However, after over a decade of data taking at different center-of-mass energies of the proton-proton collisions, the LHC has yet to find any signature of TeV scale SUSY. However, one should note that the LHC Collaborations has accumulated only ~140 fb^{-1} of data during Run-II, which is only about 5% of the planned final luminosity of 3000 fb^{-1}. Again the limits on the sparticles are mostly derived in the context of simplified SUSY models with specific assumptions on decay modes and branching ratios and the main focus has been given to prompt decays. It may very well be the case that new physics is hiding in plain sight. Our project is mainly focused on exploring the pockets of SUSY parameter space that remain untouched thus far. We plan to revisit the conventional search strategies and explore the possibilities of improving them using the new machine learning tools available to us. In the RPC SUSY electroweak sector, the constraints on masses can be significantly weaker under certain conditions, e.g., compressed scenarios. Traditional cut-based analyses fall short since the kinematics of the decay products become undistinguishable from the background. Machine learning techniques such as XGBoost and Convolutional Neural Network will be much more effective in identifying the underlying patterns and distinguishing the weak new physics scenarios. Such scenarios also have a huge impact in dark matter searches, specially for co-annihilating dark matter scenarios. Approaching from both these angles simultaneously is a more effective way to probe the unexplored parameter space and will provide with the most decisive constraints that we can derive from the existing data. For R-parity violating SUSY, we will focus on two scenarios. Large RPV coupling alongside particles with large SUSY masses and small RPV couplings with light SUSY particles that are long-lived. Similar studies in the context of RPV SUSY are seriously outdated and given the new tools and plethora of data that we have at our disposal, the results are expected to be much more impactful.