This paper explores the application of quantum-inspired machine learning, specifically using tree tensor networks (TTNs), to classify b-jets from the Large Hadron Collider (LHCb) experiment. The TTN approach is used to select important features and adapt the network geometry, optimizing for precision or speed. Results show comparable performance to deep neural networks (DNNs) but offer advantages in interpretability and adaptability for real-time applications.

Timo Felser, Marco Trenti, Lorenzo Sestini, Alessio Gianelle, Davide Zuliani, Donatella Lucchesi, Simone Montangero