Machine Learning and Econometrics in Credit Card Fraud Detection - An Empirical Analysis

Aug 2, 2026

The study addresses the rising risk of credit card fraud driven by the exponential growth of online transactions. It proposes a solution combining machine learning algorithms with econometric analysis techniques to improve detection accuracy. Using a dataset generated with PaySim to preserve privacy, the work builds and evaluates multiple machine learning models. The empirical results identify the Balanced Random Forest approach due to its accuracy and its ability to manage severe class imbalance.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages21

Reading Time53 min

AccessFree to download

Abstract

Introduction

Problem Context and Motivation

UNIVERSITA’ DEGLI STUDI DI PADOVA DIPARTIMENTO DI SCIENZE ECONOMICHE ED AZIENDALI “M. FANNO” CORSO DI LAUREA IN ECONOMIA PROVA FINALE “Machine Learning and Econometrics in Credit Card Fraud Detection: An Empirical Analysis” RELATORE: CH.MO PROF. Luca Nunziata LAUREANDO: Mattia Marzaro MATRICOLA N. 1190655 ANNO ACCADEMICO 2022 – 2023 Dichiaro di aver preso visione del “Regolamento antiplagio” approvato dal Consiglio del Dipartimento di Scienze Economiche e Aziendali e, consapevole delle conseguenze derivantida dichiarazioni mendaci, dichiaro che il presente lavoro non è già stato sottoposto, in tutto o in parte, per il conseguimento di un titolo accademico in altre Università italiane o straniere. Dichiaro inoltre che tutte le fonti utilizzate per la realizzazione del presente lavoro, inclusi imateriali digitali, sono state correttamente citate nel corpo del testo e nella sezione ‘Riferimenti bibliografici’. I hereby declare that I have read and understood the “Anti-plagiarism rules and regulations”approved by the Council of the Department of Economics and Management and Iam aware of the consequences of making false statements. I declare that this piece of work has not been previously submitted – either fully orpartially –for fulfilling the requirements of an academic degree, whether in Italy or abroad. Furthermore, I declare that the references used for this work – including the digital materials – have been appropriately cited and acknowledged in the text and in the section ‘References ’. Firma (signature) …….………………………… Abstract The exponential expansion in the volume of online transactions in the last years has evidenced how the underlying risk of credit card fraud is also rising. Addressing this topic, a potential solution can be found within the use of new computational technologies, such as machine learning algorithms, combined with econometric analysis tools. The purpose of this work is building a Machine Learning model in order to be able to detect credit card fraud with the highest degree of accuracy possible, starting from the analysis of adataset created with PaySim, a software that simulates credit card transactions based on a real dataset, due to privacy reasons. The empirical analysis was performed through diverse machine learning algorithms, resulting in the selection of a specific type of Random Forest algorithm, the Balanced Random Forest (BRT) algorithm. The selection of this specific algorithm was due to its elevated accuracy in both model building and in handling the severe class imbalance issue that emerged in the analysis of the dataset. Resulting from the analysis and model building it is safe to say that utilising new technologies as machine learning algorithms can bring a great advantage in detecting credit card fraud cases and substantially improve protection from this kind of threat that is everyday more relevant due to the progressive digitalization of our economic system. Abstract-Italiano La crescita esponenziale del volume delle transazioni online negli ultimi anni ha evidenziato come anche il rischio di frodi su carte di credito stia anch’esso aumentando. Una potenziale soluzione a questo problema può essere ricercata nelle nuove tecnologie come gli algoritmi di machine learning combinati con strumenti di analisi econometrica. Lo scopo di questo studio è la costruzione di un modello di machine learning che rilevi le frodi su carte di credito con il più alto grado di accuratezza possibile, partendo dall’analisi di un dataset creato con PaySim, un software che simula transazioni di carte di credito basandosisu un dataset reale, per motivi di privacy. L’analisi empirica è stata perseguita tramite diversi tipi di algoritmi di machine learning, con la selezione di uno specifico tipo di algoritmo, il Balanced Random Forest (BRT) . La scelta di questo specifico tipo di algoritmo è dovuta sia all’accuratezza del modello risultante che alla capacità di affront

FAQs

Why is credit card fraud detection becoming more important in digital transactions?

The volume and frequency of credit card transactions have grown rapidly, and this expansion is accompanied by an increase in reported fraud cases.

What dataset is used for building the detection model?

The analysis is based on a dataset created with PaySim, a simulator that generates credit card transactions using a real dataset while preserving privacy.

Which algorithm is selected and why?

The study selects the Balanced Random Forest (BRT) model because it delivers high accuracy and effectively handles the severe class imbalance observed in the dataset.

Aug 2, 2026

Upload by Ava Thompson