Model-Independent ML Search for New Physics

Searches for physics beyond the Standard Model often test models that are limited in scope or that are unsuited to detect unexpected discrepancies. However, recently many model-independent search algorithms have been suggested instead. These methods train neural networks (NN) under varying levels of supervision in order to discover anomalous signals in the data. In this project, we focus on a particular decay with a very well-known background, allowing us to train a signal model-independent NN that is still background model-dependent. Though the results of the project are incomplete, the framework is in place to find the minimum signal required in the data to be detectable by this method.