Publication:
Detección de transacciones fraudulentas en una plataforma de comercio electrónico

Thumbnail Image
Files

Files

Berján, David-2025.pdf (4.31 MB)
 Autorización.pdf (703.96 KB)
View Statistics
Reference Managers
Mendeley
Indexers
Google Scholar CORE
QR Code
QR Code
Authors

Authors

Berján Rodríguez, David Santiago
Abstract (Spanish)
El objetivo de la presente investigación es proponer la implementación de un algoritmo de machine learning, para determinar el nivel de riesgo de que una transacción de comercio electrónico sea fraudulenta, estableciendo el tipo fricción apropiada a cada transacción en función del grupo de riesgo en el que sea clasificada, determinando el impacto en los falsos positivos que tiene el uso de la validación de identidad. Para ello, se evaluaron diversas técnicas de preprocesamiento, incluyendo el Análisis Factorial de Datos Mixtos (FAMD), estrategias de balanceo de datos como SMOTE y ADASYN, y diferentes modelos de clasificación como regresión logística, bosques aleatorios, XGBoost y redes neuronales, los cuales se compararon entre sí.
Abstract (English)
The objective of this research is to propose the implementation of a machine learning algorithm to determine the risk level of an e-commerce transaction being fraudulent. This involves establishing the appropriate friction type for each transaction based on the risk group to which it is classified and assessing the impact of identity validation on false positives. To achieve this, various preprocessing techniques were evaluated, including Factorial Analysis of Mixed Data (FAMD), data balancing strategies such as SMOTE and ADASYN, and different classification models such as logistic regression, random forests, XGBoost, and neural networks, which were compared against each other.
Director

Advisors/Directors

Julio Niño, Andrés Felipe
Extent
57 páginas
Rights

Al consultar y hacer uso de este recurso, está aceptando las condiciones de uso establecidas por los autores.

URI: https://repositorio.escuelaing.edu.co/handle/001/3631
Collections

Collections

GE - Trabajos de Grado Maestría en Ciencia de Datos
References

21+ Chargeback Statistics You Need To Know in 2024 (2025). en. URL: https:// www.chargeback.io/blog/chargeback-statistics(visitado18-02-2025).

Badii, M. H., J. Castillo yA. Guillen (ene. de 2008). “Tamaño óptimo de la muestra”. es. En: Innovaciones de Negocios 5.9. Number: 9, págs. 53-65. ISSN: 3061-743X. DOI: 10.29105/rinn5.9-5. URL: https://revistainnovaciones.uanl. mx/index.php/revin/article/view/199 (visitado 22-04-2025).

Creswell, John W. y J. David Creswell (2023). Research Design. Qualitative, Quantitative, and Mixed Methods Approaches. en. Accepted: 2024-05-22T20:10:53Z. ISBN: 978-1-07-181796-4. URL: http://repositorio.ciem.ucr.ac.cr/ jspui/handle/123456789/514 (visitado 10-02-2025).

Dekou, Raoul et al. (2021). “Machine Learning Methods for Detecting Fraud in Online Marketplaces.” En: CIKM Workshops. URL: https://ceur-ws.org/ Vol-3052/paper15.pdf (visitado 18-02-2024).

E-commerce fraud detection prevention market 2021-2027 (2024). en. URL: https: //www.statista.com/statistics/1273278/market-size-e-commercefraud-detection-prevention-market/ (visitado 28-02-2024).

eCommerce Statistics of 2024– Forbes Advisor (2024). URL: https://www. forbes.com/advisor/business/ecommerce-statistics/ (visitado 27-02-2024).

Gupta, Surbhi, N Malsa y Mr Vimal Gupta (2017). “Credit card fraud detection and prevention—a survey”. En: International Journal for Innovative Research in Science & Technology 4, págs. 1-7.

Hou, Chun Kit Jeffery y Kamran Behdinan (dic. de 2022). “Dimensionality Reduction in Surrogate Modeling: A Review of Combined Methods”. en. En: Data Science and Engineering 7.4, págs. 402-427. ISSN: 2364-1541. DOI: 10.1007/ s41019-022-00193-5. URL: https://doi.org/10.1007/s41019-02200193-5 (visitado 21-12-2024).

Levine, E. S. (2012). “Improving risk matrices: the advantages of logarithmically scaled axes”. en. En: Journal of Risk Research 15.2. Publisher: Taylor & Francis Journals, págs. 209-222. URL: https://ideas.repec.org//a/taf/ jriskr/v15y2012i2p209-222.html (visitado 27-02-2025).

M, Hossin y Sulaiman M.N (mar. de 2015). “AReview on Evaluation Metrics for Data Classification Evaluations”. en. En: International Journal of Data Mining &KnowledgeManagementProcess5.2,págs.01-11. ISSN: 2231007X, 22309608. DOI: 10.5121/ijdkp.2015.5201. URL: http://www.aircconline.com/ ijdkp/V5N2/5215ijdkp01.pdf (visitado 22-01-2025).

Mohsen, Omar Rajab, Ghalia Nassreddine y Mazen Massoud (4 de jul. de 2023). “Credit Card Fraud Detector Based on Machine Learning Techniques”. En: Journal of Computer Science and Technology Studies 5.2. Number: 2, págs. 16-30. ISSN: 2709-104X. DOI: 10.32996/jcsts.2023.5.2.2. URL: https:// www.al-kindipublisher.com/index.php/jcsts/article/view/5515 (visitado 18-02-2024).

Ogunleye, Adeola y Qing-Guo Wang (nov. de 2020). “XGBoost Model for Chronic Kidney Disease Diagnosis”. En: IEEE/ACM Transactions on Computational Biology and Bioinformatics 17.6. Conference Name: IEEE/ACM Transactions on Computational Biology and Bioinformatics, págs. 2131-2140. ISSN: 15579964. DOI: 10.1109/TCBB.2019.2911071. URL: https://ieeexplore. ieee.org/abstract/document/8693581 (visitado 22-01-2025).

Saputra, Adi y Suharjito (2019). “Fraud Detection using Machine Learning in e-Commerce”. En: Publisher: Science and Information (SAI) Organization Limited. ISSN: 2158107X. DOI: 10.14569/IJACSA.2019.0100943. URL: https://www.proquest.com/docview/2655164322?pq-origsite= gscholar&fromopenview=true&sourcetype=Scholarly%20Journals (visitado 18-02-2024).

Visa Secure, EMV 3-D Secure para comercios (2024). es. URL: https://www. visa.com.co/run-your-business/small-business-tools/paymenttechnology/visa-secure.html (visitado 30-03-2024).

Xuan, Shiyang et al. (mar. de 2018). “Random forest for credit card fraud detection”. En: 2018 IEEE 15th International Conference on Networking, Sensing and Control (ICNSC), págs. 1-6. DOI: 10.1109/ICNSC.2018.8361343. URL: https://ieeexplore.ieee.org/abstract/document/8361343 (visitado 15-01-2025).

Zhang, Ge et al. (7 de mar. de 2022). “eFraudCom:An E-commerce Fraud Detection System via Competitive Graph Neural Networks”. En: ACM Transactions on Information Systems 40.3, 47:1-47:29. ISSN: 1046-8188. DOI: 10.1145/3474379. URL: https://doi.org/10.1145/3474379 (visitado 18-02-2024).