A Proposed Model for Card Fraud Detection Based on CatBoost and Deep Neural Network

Abstract

The rapid development of technology has digitized customer payment behavior towards a cashless society. To a certain extent, this has created a feast for miscreants to commit fraud. According to Nilson (2020), global fraud loss is projected to reach over $35 billion by 2025. Consequently, the need for a novel method to prevent this menace is undisputed. This research was conducted on the IEEE-CIS Fraud Detection Dataset provided by Vesta Corporation. Based on the logic of labeling for converting the entire account to ''Fraud=1'' once the credit card has fraud, we navigate the research process towards predicting fraudulent credit cards rather than fraudulent transactions. The key idea behind the proposed model is user separation, in which we divide users into old and new people before applying CatBoost and Deep Neural Network to each category, respectively. In addition, a variety of techniques to improve detection accuracy, namely handling heavily imbalanced datasets, feature transformation, and feature engineering, are also presented in detail in this paper. The experimental results showed that our model performed well, as we obtained AUC scores of 0.97 (CatBoost) and 0.84 (Deep Neural Network).

Arquitetura de Tempo Real e Modelo de Aprendizado de Máquina para Detecção de Fraudes de Cartão de Crédito

“…Na Tabela 2, é possível observar os resultados apresentados pelo melhor modelo encontrado pela abordagem proposta para cada um dos cinco algoritmos de classificac ¸ão investigados (Logistic Regression, Decision Tree (DT), K-Nearest Neighbors (KNN), Random Forest (RF) e XGBoost, bem como os resultados reportados pelos principais trabalhos relacionados: [Deng et al 2021], [Kewei et al 2021], [Nguyen et al 2022], [Chen e Han 2021], [Ni et al 2023] e [Bakhtiari, Nasiri e Vahidi 2023]. Vale destacar que todos esses trabalhos utilizaram o conjunto de dados IEE-CIS Fraud Detection.…”

“…O estudo em [Nguyen et al 2022] apresenta um modelo híbrido usando CatBoost e Rede Neural Profunda para detectar fraudes, distinguindo entre usuários novos e antigos. O escopo do trabalho se concentrou na etapa de pré-processamento e na engenharia de atributos, com a atenc ¸ão voltada para os dados pertinentes às transac ¸ões, como valor, data e horário, provenientes de clientes já conhecidos, além dos atributos ligados aos cartões, como tipo e país, quando aplicáveis aos novos usuários.…”

Dark Web Financial Fraud Identification Using Mathematical Models in Healthcare Domain

“…Nguyen et al [5] recommend several different strategies that can be used to detect fraudulent behavior more effectively. It focuses on enhancing detection skills by modifying existing characteristics and establishing new ones, as well as coping with datasets with significant variation.…”

“…Next, hospitals will be grouped hierarchically [22]. First, second, and third clusters contain medical facilities in order (5,6,7,9,10,11). Hospitals were grouped into the four groups shown.…”

Credit Card Fraud Detection: An Improved Strategy for High Recall Using KNN, LDA, and Linear Regression

“…Nguyen et al [ 29 ] devised an advanced framework for real-time credit card fraud detection. Prior to feeding data into the deep learning model, they implemented a distinction mechanism to classify credit card users as either longstanding or newcomers.…”