Makine öğrenimi teknikleriyle alt kuark kaynaklı jet etiketlemesinin incelenmesi

In this study, Machine Learning methods are applied to the problem of classifying b quark-originated jets in particle physics. For this purpose, a custom dataset was specifically prepared, and Multi-Layer Perceptron (MLP), XGBoost, and Retentive Networks models were tested. Among these, the Retentive Network-based architecture, JetRetNet, demonstrated superior performance. Performance evaluations were conducted by comparing JetRetNet with state-of-the-art models such as DeepJet and Particle Transformer. While JetRetNet does not outperform these models, it shows significant potential for future applications in b-jet tagging with further data and optimization.