Thesis Defense – Ayşe Asu Güvenli (MSPHYS)

Ayşe Asu Güvenli – M.Sc. in Physics

Prof. Dr. Taylan Akdoğan – Advisor

Prof. Dr. Bora Işıldak - Co - Adviosr

Date: 27.12.2024

Time: 13.00

Location: AB1 412

“EXPLORATION OF BOTTOM QUARK ORIGINATED JET TAGGING WITH MACHINE LEARNING TECHNIQUES”

Prof. Dr. Taylan Akdoğan, Özyeğin University

Prof. Dr. Bora Işıldak, Yıldız Technical University

Prof. Mehmet Burçin Ünlü, Özyeğin University

Doç Dr. Kaan Güven, Özyeğin University

Dr. Öğrt. Üyesi Parviz Elahi, Özyeğin University

Abstract:

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 specif- ically 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 com- paring 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 fu- ture applications in b-jet tagging with further data and optimization.

Bio:

Ayşe Asu Güvenli received her B.Sc degree in Physics from Boğaziçi University. She is currently pursuing an M.Sc in Physics from Özyeğin University, and her research interests are in particle physics and the application of machine learning techniques to physics problems.