Boosting LHC data with machine learning

The LHC experiments produce vast amounts of data. The analysis chain to produce physics results from these data is a prime candidate for modern machine learning applications and deep learning in particular with the goal to maximise the information we extract, to speed up the whole analysis chain and to simplify procedures. Examples include deciding which data to record (triggering) and making it available for analysis, guaranteeing highest quality of the data, reconstruction of high-fidelity physics objects, exquisite simulation of the detector response, analysis optimisation, treatment of systematic uncertainties, interpretation of results, searching for anomalies in the data, and the design of new detectors. In the seminar the most interesting and promising examples will be highlighted.