Collider signatures for Hidden Valley jets

Abstract

This thesis delves into the realm of Hidden Valley models, which introduce the intriguing concept of a new gauge group alongside the Standard Model gauge group, connected by a heavy mediator particle. Seeking to discern their signatures at the Large Hadron Collider (LHC), a binary classification MLP (Multi-Layer Perceptron) neural network serves as a key tool for signal discrimination. The introduction of the SIFT algorithm enhances the extraction of dark meson mass scales and discriminating the Hidden Valley jets against ordinary quark or gluon jets. We also introduce other benchmarks to explore sensitivity to heavy and light quark decays of Hidden Valley mesons. The use CNNs, with jet-images as inputs, offers an alternative avenue to study the properties of Hidden Valley jets. This comprehensive analysis unveils the potential to decipher elusive signals of the Hidden Valley.