Data collected by the LHC experiments is valuable and finite. At the same time, the physics processes which are being searched for or measured, and the corresponding backgrounds predicted by the standard model have non-trivial and multi-dimensional underlying kinematics of the produced particles, and the space of observables is even further expanded by the interaction with and measurement by the detectors. This combined with the availability of large control samples from simulation or LHC data itself provides many opportunities for the use of machine learning to extend the sensitivity of searches and improve the precision of measurements at the LHC. I'll discuss the usage of machine learning techniques at the LHC experiments, covering both the reconstruction and identification of physics objects as well as usage in higher level physics analyses, with a particular focus on the discovery and measurements of the Higgs Boson in CMS. I'll also discuss possible uses of machine learning for numerical phase-space integration in Monte Carlo generators.
