Anomaly Awareness

We will exemplify the use of this method in a nontrivial Algorithms that detect anomalies have to learn normal task in our field-domain, Particle Physics. In the context behaviour to be able to identify anomalous behaviour. of the Large Hadron Collider (LHC) searches for new Sometimes we do know what types of anomalies we need phenomena, we show how Anomaly Awareness can help to search for, and then use supervised Machine Learning making these searches more robust, less dependent on the (ML) methods to find them. As anomalies are, by definition, specific scenarios one has in mind. This model-independence rarer than normal events, these supervised techniques need to of LHC searches is particularly important now that be adapted to unbalanced datasets and made robust against the traditional ways of thinking in Particle Physics are fluctuations in the dominant normal or in-distribution dataset.