Credit Risk Models for Financial Fraud Detection

Abstract

Outlier detection is currently applied in many fields, where existing research focuses on improving imbalanced data or enhancing classification accuracy. In the financial area, financial fraud detection puts higher demands on real-time and interpretability. This paper attempts to develop a credit risk model for financial fraud detection based on an extreme gradient boosting tree (XGBoost). SMOTE is adopted to deal with imbalanced data. AUC is the assessment indicator, and the running time is taken as the reference to compare with other frequently used classification algorithms. The results indicate that the method proposed by this paper performs better than others. At the same time, XGBoost can obtain a ranking of important features that impact the classification results when performing classification tasks, making the evaluation results of the model interpretable. The above shows that the model proposed in the paper is more practical in solving credit risk assessment problems. It has faster response times, reduced costs, and better interpretability.

Handling Imbalanced Data With Weighted Logistic Regression and Propensity Score Matching methods

“…Data imbalance refers to an asymmetrical distribution of data across different classes or categories within a dataset, whereby certain classes are underrepresented compared to others (He & Garcia, 2009). This issue can manifest in diverse contexts, including financial fraud detection (Xia & Zhang, 2023;Al-Shabi, 2019), medical diagnosis (Bridge et al, 2020), sentiment analysis (Al Shamsi & Abdallah, 2022), text analysis (Li et al, 2023), image classification (Tian & Han, 2022), and recommendation systems (Zhang et al, 2019).…”

Cluster-Based Oversampling with Area Extraction from Representative Points for Class Imbalance Learning

“…In SMOTE, additional minority samples are created along the line segment among the minority samples, although with no indication of any kind to the samples available in the confrontational majority class. SMOTE has been applied in various domains, including finance (Sun et al, 2020), fraud detection (Xia et al, 2023), medical diagnosis (Bokhare et al, 2023, Kamarulzalis et al, 2018 and image classification (Khan & Sheikh, 2023). It has shown promising results in improving the classification accuracy of models in these domains.…”