A hybrid semi-supervised approach for financial fraud detection

Liu, Jin-Miao; Tian, Jiang; Cai, Zhu-Xi; Zhou, Yongqin; Luo, Ren-Hua; Wang, Ranran

doi:10.1109/icmlc.2017.8107767

“…The most commonly used algorithm for credit evaluation in banking system is logistic regression [13, 14]. Besides, decision tree is a well-known algorithm to predict individuals' credit, such as credit card fraud [15]. In spite of remarkable accuracy and simple construction of the algorithm, a credit model based on logistic regression method possesses strong interpretability, which is indeed favorable in banking system [16].…”

Section: Background Of Credit Evaluation Methods and P2p Lending Smentioning

confidence: 99%

A Trial of Student Self-Sponsored Peer-to-Peer Lending Based on Credit Evaluation Using Big Data Analysis

Hou

¹

,

Ma

²

,

Mei

³

et al. 2019

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

There is still no effective approach to overcome the problem of credit evaluation for Chinese students. In absence of a reliable credit evaluation system for students, the university students have to only apply through online peer-to-peer (P2P) loan platforms because Chinese financial institutions typically reject students’ loan applications. Lack of students’ financial records hinders financial institutes and banks to routinely evaluate the students’ credit status and assign loans to them. Hence, this paper attempted to benefit from university students’ diversified daily behavior data, and logistic regression (LR) and gradient boosting decision tree (GBDT) algorithms were also used to develop robust credit evaluation models for university students, in which the validation of the proposed models was assessed by a real-time P2P lending platform. In this study, the students’ overdue behavior in returning books to university library was used as an index. With training 17838 samples, the proposed models performed well, while GBDT-based model outperformed in identification of “bad borrowers.” Based on the proposed models, a self-sponsored peer-to-peer loan platform was established and developed in a Chinese university for ten months, and the achieved findings demonstrated that adopting such credit evaluation models can effectively reduce the default ratio.

show abstract

“…Model-based approaches such as [9], [10] and [11] are designed to improve the limitations of standard supervised and unsupervised approaches to anomaly detection. They do this by introducing new approaches and creating combinations and/or variants of existing implementations.…”

Section: Model-based Approachesmentioning

confidence: 99%

Combining Artificial Immune System and Clustering Analysis: A Stock Market Anomaly Detection Model

Close

2023

Preprint

0

View full text Add to dashboard Cite

<p>As online trading systems have increased the amount of high volume, real-time data transactions, the stock market continues to have increased vulnerability to attacks. The goal of this project is to detect these attacks based on normal trade behavior using an Artificial Immune System (AIS) approach combined with clustering algorithms. The AIS approach is inspired by its proven ability to handle time-series data as well as its ability to detect abnormal behavior while only being trained on regular trade behavior. These two main points are important as the models need to be able to adapt over time to adjust to normal trade behavior as it evolves and also due to confidentiality and data restrictions, real-world manipulations are not available for training. This proposed research work discovers a competitive alternative to the leading AIS approach and investigates the effects of combining AIS with clustering algorithms including Kernel Density Estimation, Self-Organized Maps, Density-Based Spatial Clustering of Applications with Noise, Cluster-Based Local Outlier Factor, and Spectral clustering. The best performing solution achieves leading performance in terms of many common clustering metrics including Area Under the Curve, False Alarm Rate, False Negative Rate, and Computation Time. </p>

show abstract

“…It resulted in an AUC value of over 0.8 for all datasets it was tested on Although effective, this approach used a limited dataset only using ask and bid price information and is also computationally complex. A hybrid semi-supervised model is used in [11] to detect fraud in bank wire transfers. The hybrid combination outperformed clustering techniques by reducing the false positive rate and showed it can handle real time data well.…”

Section: Model-based Approachesmentioning

confidence: 99%

“…However, current solutions that use unsupervised learning methods [4], [8][9][10][11][12][13][14] have suffered from problems such as reliance on parameter tweaking, representing time series data effectively and class imbalance. In [2], a semi-supervised approach is introduced which mimics the human immune system called Dendritic Cell Algorithm (DCA).…”

Section: Chapter 1: Introductionmentioning

confidence: 99%

Combining Artificial Immune System and Clustering Analysis: A Stock Market Anomaly Detection Model

Close

2023

Preprint

0

View full text Add to dashboard Cite

<p>As online trading systems have increased the amount of high volume, real-time data transactions, the stock market continues to have increased vulnerability to attacks. The goal of this project is to detect these attacks based on normal trade behavior using an Artificial Immune System (AIS) approach combined with clustering algorithms. The AIS approach is inspired by its proven ability to handle time-series data as well as its ability to detect abnormal behavior while only being trained on regular trade behavior. These two main points are important as the models need to be able to adapt over time to adjust to normal trade behavior as it evolves and also due to confidentiality and data restrictions, real-world manipulations are not available for training. This proposed research work discovers a competitive alternative to the leading AIS approach and investigates the effects of combining AIS with clustering algorithms including Kernel Density Estimation, Self-Organized Maps, Density-Based Spatial Clustering of Applications with Noise, Cluster-Based Local Outlier Factor, and Spectral clustering. The best performing solution achieves leading performance in terms of many common clustering metrics including Area Under the Curve, False Alarm Rate, False Negative Rate, and Computation Time. </p>

show abstract

A hybrid semi-supervised approach for financial fraud detection

Cited by 20 publications

References 5 publications

A Trial of Student Self-Sponsored Peer-to-Peer Lending Based on Credit Evaluation Using Big Data Analysis

A Trial of Student Self-Sponsored Peer-to-Peer Lending Based on Credit Evaluation Using Big Data Analysis

Combining Artificial Immune System and Clustering Analysis: A Stock Market Anomaly Detection Model

Combining Artificial Immune System and Clustering Analysis: A Stock Market Anomaly Detection Model

Contact Info

Product

Resources

About