Exploring New Physics Through Collider and Gravitational Wave Measurements with Artificial Neural Networks: the Case Study of QCD-like Technicolor

Abstract: With physicists actively exploring Beyond the Standard Model (BSM) theories that may fill in the gaps of the Standard Model (SM), a fundamental question arises: which parameters hold physical significance? In this thesis, we present our initial progress towards the development of a model-independent artificial intelligence framework designed for conducting parameter space scans in BSM scenarios. Our framework incorporates several publicly available high-energy physics packages, namely SPheno, HiggsBounds, HiggsSignals, and CosmoTransitions. These packages enable us to impose various constraints including unitarity, electroweak precision, Higgs exclusion limits, and strong detectable first-order cosmic phase transitions. To demonstrate the effectiveness of our framework, we apply it to a singlet-triplet extended SM, serving as a minimal low-scale effective field theory for a quantum chromodynamic-like Technicolor (TC) theory. A proper phenomenological investigation and parameter space analysis of the TC theory in the UV-limit are planned for future work. The findings from our preliminary investigation exhibit promising results, demonstrating a substantial efficiency enhancement when compared to conventional random search approaches for identifying physically relevant parameter points. These outcomes pave the way for future BSM studies utilizing our developed framework.