Is the machine smarter than the theorist: Deriving formulas for particle kinematics with symbolic regression

We demonstrate the use of symbolic regression in deriving analytical formulas, which are needed at various stages a typical experimental analysis collider phenomenology. As first application, we consider kinematic variables like stransverse mass, ${M}_{T2}$, defined algorithmically through an optimization procedure and not terms formula. then train obtain correct expressions for all known special cases ${M}_{T2}$ literature. second reproduce expression next-to-leading order (NLO) distribution from data, is simulated with NLO event generator. Finally, derive approximations distributions after detector simulation, no formulas currently exist.