Thesis Defence – Sencer Kaya

Thesis Defense - Sencer Kaya

Assistant Professor Dr. Levent Güntay - Advisor

Date: 27th July 2026

Time: 16:00

Location: Özyeğin University Altunizade Campus - Classroom ALT 101

“Quantum Machine Learning for Credit Card Fraud Detection: A Systematic Comparison”

Assistant Professor Dr. Levent Güntay, Özyeğin University - Advisor

Assistant Professor Dr. Emrah Ahi, Özyeğin University

Professor Dr. Adnan Fatih Kocamaz, İnönü University

Abstract:

The detection of credit card fraud presents a complex classification challenge due to the significant imbalance between fraudulent and legitimate transactions. This thesis aims to examine the potential of quantum machine learning in offering insightful representations for this issue by systematically comparing three key approaches: variational quantum classifiers, quantum reservoir computing, and quantum kernel methods. The original transaction data undergo a reduction to qubit-level feature sets employing Neighbourhood Components Analysis, and quantum models are assessed in conjunction with well-established classical benchmarks within a secure training, validation, and testing framework. The study delves into the impact of data encoding, circuit architecture, dimensionality, and encoding magnitude, while judging performance through metrics suited to highly imbalanced data distributions. The results demonstrate that meticulously crafted quantum configurations can stand as strong contenders with, and under specific conditions, surpass classical models. The research also pinpoints encoding magnitude as a pivotal element in governing quantum-kernel concentration and model efficacy. On the whole, the thesis delivers a comprehensive empirical evaluation of the circumstances in which quantum learning methodologies could prove beneficial in the domain of credit card fraud detection.

Keywords: quantum machine learning, credit card fraud detection, variational quantum classifiers, quantum reservoir computing, quantum kernel methods

Bio:

Sencer Kaya holds a bachelor's degree in physics and is currently pursuing a master's degree in Financial Engineering at Özyeğin University. He is employed as a Quantum Software Engineer at Q&Co, with research interests spanning quantum machine learning, quantum optimization, financial engineering, and the utilization of quantum computing in practical decision-making scenarios.