Transaction Dataset Research Articles

Real-time Detection of Abnormal Financial Transactions Using Generative Adversarial Networks: An Enterprise Application

This paper presents a novel real-time financial fraud detection framework utilizing Generative Adversarial Networks (GANs) for enterprise applications. The proposed system addresses critical challenges in fraud detection, including class imbalance, real-time processing requirements, and enterprise scalability. Implementing a sophisticated multi-layered architecture, the system integrates advanced preprocessing techniques with an optimized GAN model explicitly designed for fraud pattern recognition. The framework incorporates parallel processing capabilities and adaptive batch processing mechanisms to maintain high throughput while ensuring sub-second latency. The experimental evaluation uses a subset of the European Credit Card Transaction dataset, comprising 50,000 transactions with a balanced representation achieved through strategic sampling and SMOTE technique. The proposed model achieves 97.8% accuracy, 96.5% precision, and 95.8% recall, demonstrating competitive performance compared to traditional machine learning approaches. Real-time performance analysis shows consistent sub-100ms latency while maintaining robust performance under varying load conditions. The system demonstrates linear scalability up to 32 nodes, with high availability and failover capabilities. The comprehensive assessment validates the framework's effectiveness in enterprise environments, providing practical solutions for financial institutions facing evolving fraud challenges. This research contributes to the advancement of financial security technology through the innovative application of adversarial learning in fraud detection.

Cross-shard transaction optimization based on community detection in sharding blockchain systems

Blockchain systems have always faced the challenge of performance bottlenecks, and sharding technology is considered a promising mainstream on-chain scalability solution to solve this problem. Due to the complexity and high cost of the cross-shard transaction processing mechanism in the sharding blockchain system, as well as the high proportion of cross-shard transactions, it becomes challenging for the sharding blockchain system to reach the ideal theoretical performance upper limit. Therefore, this paper aims to reduce the proportion of cross-shard transactions by dividing accounts with frequent transactions into the same shard, thereby improving system throughput. This paper builds a hypergraph based on historical transaction data to represent the diverse transaction relationships between accounts, and formulates the account division problem in the blockchain as a community discovery problem on the hypergraph structure. A time-aware community detection algorithm is proposed to partition accounts by considering the sustainability of transaction relationships between accounts. This also solves the problem of community detection algorithms tending to partition into larger shards. In addition, this paper builds a local Ethereum test network and implements the proposed algorithm on a real transaction dataset. Experimental results show that this algorithm can reduce the proportion of cross-shard transactions from about 95% to about 10%. Furthermore, it shows superior performance in terms of transaction throughput and latency compared with other community detection-based account partitioning algorithms.

Deep Learning-Based Freight Recommendation System for Freight Brokerage Platform

Platform-based businesses in the logistics market are evolving under the influence of digital transformation. Transforming the freight market into an environment where various types of freight can be traded across multiple markets and locations. Freight brokerage platforms have revolutionized the trading relationship between freight owners and vehicle owners. However, this type of system has also introduced inefficiencies, such as unestablished contracts, leading to unnecessary costs and delays. To address this issue, a freight recommendation system can assist users in finding what they are looking for while aiming to reduce failed contracts. With current advances in deep learning, complex patterns based on users’ past behaviors and preferences can be learned, enabling more accurate and personalized recommendations. This study proposes a deep learning-based freight recommendation system to provide personalized services and reduce failed contracts on freight brokerage platforms. The system is built by creating a freight transaction dataset, classifying freight categories through natural language processing and text mining techniques, and incorporating externally derived data on transportation distances. The deep learning model is trained using Autoencoder, Word2Vec, and Graph Neural Networks (GNN), with recommendation logic implemented to suggest suitable freight matches for vehicle owners. This system is expected to increase the market efficiency of the freight logistics industry and is a key step toward improving the long-term profit structure.

TRANSFORMING BANKING SECURITY: THE ROLE OF DEEP LEARNING IN FRAUD DETECTION SYSTEMS

In the digital banking landscape, the increasing volume of online transactions has heightened the risk of fraudulent activities, necessitating the development of more effective detection systems. This study investigates the efficacy of various machine learning and deep learning algorithms in identifying fraudulent transactions, emphasizing Long Short-Term Memory (LSTM) networks. We implemented and evaluated multiple algorithms, including Logistic Regression, Random Forest, Gradient Boosting Machines (GBM), and XGBoost, on a large-scale credit card transaction dataset. Our results demonstrate that the LSTM model outperforms traditional machine learning algorithms, achieving an accuracy of 98.5%, precision of 87.2%, recall of 85.0%, and an Area Under the Curve (AUC) score of 0.94. These findings highlight the superior capability of LSTM networks to capture complex patterns in sequential transaction data, making them an asset for real-time fraud detection in banking. This research underscores the need for financial institutions to adopt advanced deep learning techniques to enhance their fraud detection systems, thereby minimizing financial losses and improving customer trust.

The effect of subsidized flood insurance on real estate markets

AbstractSubsidized insurance against extreme weather events improves affordability among households in at‐risk areas but it can weaken the risk signal via property prices. Leveraging a granular data set of all property transactions and flood events in England, we study the effects of a reinsurance scheme that lowers insurance premiums for at‐risk properties. We document that the introduction of this scheme increases prices and transaction volumes of flood‐prone properties. This fully offsets the negative direct effects of flooding on property prices, with high‐income areas and high‐value properties benefiting relatively more. Our findings speak to the debate on climate adaptation policies and their consequences for wealth distribution.

Advancing Fraud Detection in Banking: Integration of Data Pipelines, Machine Learning, and Cloud Computing

This article examines the transformation of fraud detection in the banking sector through the integration of advanced data pipelines, machine learning (ML), artificial intelligence (AI), and cloud computing technologies. We analyze how modern data pipelines enable real-time processing of vast transactional datasets, significantly improving the timeliness and accuracy of fraud detection. The article explores the application of ML models in identifying suspicious patterns and the role of AI-driven systems in continuously adapting to evolving fraud schemes while reducing false positives. We evaluate the impact of cloud platforms such as AWS, Azure, and Google Cloud in providing scalable, cost-efficient infrastructures for processing massive datasets and supporting seamless integration with ML models. The article presents case studies of successful implementations by major banks, demonstrating substantial reductions in processing times and improved detection efficiency. Additionally, we address key challenges including data quality maintenance, model interpretability, and false positive mitigation. The article concludes by discussing future innovations such as Federated Learning and Explainable AI (XAI), which promise to enhance cross-institutional collaboration and decision-making transparency in fraud detection. This comprehensive analysis provides valuable insights for financial institutions seeking to enhance their fraud detection capabilities in an increasingly complex digital landscape.

Impact of waterborne E. coli outbreaks on local communities: evidence from housing transactions in Michigan

Abstract Waterborne Escherichia coli (E. coli) represents a pervasive water quality problem across the United States. In Michigan, the presence of E. coli has become problematic for many areas where agricultural run-off and ineffective policies have made these outbreaks endemic. Combining the universe of housing transaction datasets from 2009 to 2017 with the State of Michigan water sampling dataset, we investigate and quantify the negative impacts of E. coli outbreaks on local housing prices. Our difference-in-differences model estimates an overall impact of −8.94% for houses in the treatment group relative to the control group. However, this effect is only short term, as sales prices recover after the outbreak has ended.

Applying Bayesian Networks for Detecting Bank Fraudulent Transactions in Nigeria

Bank fraud is an increasingly prevalent issue, causing substantial financial losses for both financial institutions and customers. Also, with the increasing prevalence of bank fraud, especially facilitated by online banking and digital payments, there is a pressing need for potential expertise deployment and advanced fraud detection techniques in real-world banking systems. This study specified the Bayesian Network models for detecting fraudulent bank transactions. Bayesian Networks offer a probabilistic graphical modeling approach that can effectively capture complex relationships and dependencies within financial data. Thus, the research aimed at developing a custom Bayesian network model trained on a large dataset of bank transactions comprising over one million bank transactions to classify instances as fraudulent or non-fraudulent. Python 3.11.10 was used for the graphical representations. Down-sampling was employed to reduce the dataset from its initial size of 1 million observations to 17,650 observations, ensuring a balanced representation of both fraud and non-fraud instances. Various estimation techniques: Maximum Likelihood, Bayesian Networks, and Expectation Maximization were employed and evaluated to learn the model parameters. The model's performance was measured using metrics: accuracy, precision, recall, F1-score, and ROC-AUC. The Bayesian Networks estimator achieved an overall accuracy of 66.18%, precision of 67.73%, F1-score of 64.06%, ROC-AUC of 66.13%, Recall of 60.77%, Sensitivity of 60.77% and Specificity of 71.50%. Also, the Maximum Likelihood achieved an overall Accuracy of 66.77%, Precision of 69.22%, F1-score of 63.96%, ROC-AUC of 66.71%, Recall of 59.45%, Sensitivity of 59.45% and Specificity of 73.97%. Likewise, the Expectation Maximization achieved an overall Accuracy of 66.83%, Precision of 69.31%, F1-score of 64.00%, ROC-AUC of 66.77%, Recall of 59.45%, Sensitivity of 59.45% and Specificity of 74.09%. On the confusion matrix, the model correctly classified 1272 instances as Non-Fraudulent transactions (True Negative). Also, it was observed that the model incorrectly classified 507 instances as fraudulent transactions when they were Non-Fraudulent (False Positive). The model similarly incorrectly classified 687 instances as Non-Fraudulent transactions when they were Fraudulent (False Negative). Finally, it correctly classified 1064 instances as fraudulent transactions (True Positive). These results demonstrated the Bayesian Network's ability to identify fraudulent transactions while minimizing false alarms accurately. The findings highlight the potential of Bayesian Networks as a robust framework for fraud detection in the banking sector, contributing to enhanced security and reduced financial losses. The developed model can be introduced into existing banking systems to strengthen fraud prevention strategies.

Do non-farmers pay more for land than farmers?

Abstract The increase in farmland prices in many parts of the world over the past decade has sparked discussions about whether non-farmers pay higher prices for farmland. This study uses a causal machine learning approach with a rich data set of land transactions in Germany to quantify the potential price premium paid by non-farmers. By applying the causal forest method, we uncover the heterogeneity of price premiums and reveal moderating effects of covariates. We find that the average positive price premium by non-farmers decreases with parcel size and distance to a highway exit.

Digital Twin and IoT-Enabled Online Learning Model for Real- Time Monitoring and Control of Fraudulent Transaction in the Banking industry

Fraudulent transactions have long been a challenge in the banking industry, causing significant financial losses and operational burdens. Traditional fraud detection systems often struggle to adapt and respond in real-time to the evolving tactics of fraud. This study aims to develop a digital twin and IoT-enabled online learning model for real-time monitoring and control of fraudulent transactions in banking. The proposed system integrates digital twin technology and IoT sensors to create a virtual replica of transaction environments, enabling continuous learning and adaptation using an online learning model. Performance metrics such as detection speed, prediction accuracy, precision, recall, and F1 score were used to evaluate the system. The system was tested using a dataset of banking transactions, including legitimate and fraudulent activities. Key results include detecting fraudulent transactions 40% faster than traditional methods, reducing detection time from 5 minutes to 3 minutes. The model achieved an accuracy of 92%, an 8.24% improvement over traditional fraud detection systems. The false positive rate was reduced by 28%, meaning fewer legitimate transactions were wrongly flagged as fraudulent. Precision increased by 5.95%, reaching 89%, and recall improved by 8.75%, reaching 87%. This led to a 7.32% increase in the F1 score, reaching 88%. The system offers enhanced real-time fraud detection capabilities, reducing false positives and improving overall security and operational efficiency in the banking industry.

DEEP LEARNING MODEL FOR DETECTING TERROR FINANCING PATTERNS IN FINANCIAL TRANSACTIONS

Terror financing remains a critical threat to global security, with illicit actors continually adapting their methods to evade detection. Traditional financial monitoring systems often struggle to identify the complex and covert patterns associated with terror-related transactions due to their reliance on predefined rules and statistical thresholds. This study introduces advanced deep learning models designed to detect terror financing patterns within vast datasets of financial transactions. We developed and evaluated several neural network architectures, including Convolutional Neural Networks (CNNs) and Long Short-Term Memory networks (LSTMs), to capture both spatial and temporal transaction features. The models were trained and tested on a dataset comprising anonymized financial transactions labeled for suspicious activities related to terror financing. Our deep learning models demonstrated superior performance over conventional machine learning approaches, achieving higher accuracy, precision, and recall in identifying suspicious transactions. Notably, the LSTM-based model excelled in detecting sequential transaction patterns indicative of layering and integration stages commonly used in terror financing. The results underscore the potential of deep learning techniques in enhancing the capabilities of financial institutions and regulatory bodies to combat terror financing. Implementing such models can lead to more proactive and effective monitoring systems that adapt to evolving illicit financing strategies.

PREDICTIVE ANALYSIS OF CRYPTOCURRENCY INCOME USING MACHINE LEARNING MODELS

This research explores the potential of various machine learning models to predict cryptocurrency income. With the increasing volatility and widespread use of cryptocurrencies, accurately forecasting income from these digital assets is crucial for investors and stakeholders. By examining a dataset of cryptocurrency transactions, this study addresses the challenge of outliers and assesses the correlations between different variables to ensure robust data analysis. We employ three distinct machine learning models—Linear Regression, Decision Tree Regressor, and Random Forest Regressor—to train and evaluate the dataset. Each model is analysed to calculate performance metrics, such as Mean Squared Error (MSE) and R2 Score, which provide insights into their predictive accuracy and reliability. The detailed analysis and discussion highlight the strengths and weaknesses of each approach, emphasizing the efficiency and accuracy of the models in forecasting cryptocurrency income. The Linear Regression model, known for its simplicity, is compared against more complex models like Decision Tree and Random Forest Regressors, which can capture nonlinear relationships and interactions between variables. The Random Forest Regressor, an ensemble learning method, is particularly noted for its superior performance in handling complex datasets and providing more accurate predictions. This study's findings offer valuable insights for developing more effective investment strategies and risk management practices in the rapidly evolving cryptocurrency market. Through this comprehensive evaluation, the research aims to contribute to the growing body of knowledge on financial forecasting using machine learning techniques, paving the way for future advancements in this field.

Non-Medical Switching or Discontinuation Patterns among Patients with Non-Valvular Atrial Fibrillation Treated with Direct Oral Anticoagulants in the United States: A Claims-Based Analysis.

This study assessed direct-acting oral anticoagulant (DOAC) switching/discontinuation patterns in patients with non-valvular atrial fibrillation (NVAF) in 2019, by quarter (Q1-Q4), and associated socioeconomic risk factors. Adults with NVAF initiating stable DOAC treatment (July 2018-December 2018) were selected from Symphony Health Solutions' Patient Transactional Datasets (April 2017-January 2021). Switching/discontinuation rates were reported in 2019 Q1-Q4, separately. Non-medical switching/discontinuation (NMSD) was defined as the difference between switching/discontinuation rates in Q1 and mean rates across Q2-Q4. The associations of socioeconomic factors with switching/discontinuation were assessed. Of 46,793 patients (78.7% ≥ 65 years; 52.6% male; 7.7% Black), 18.0% switched/discontinued their initial DOAC in Q1 vs. 8.8% on average in Q2-Q4, corresponding to an NMSD of 9.2%. During the quarter following the switch/discontinuation, more patients who switched/discontinued in Q1 remained untreated (Q1: 77.0%; Q2: 74.3%; Q3: 71.2%) and fewer reinitiated initial DOAC (Q1: 17.6%; Q2: 20.8%; Q3: 24.0%). Factors associated with the risk of switching/discontinuation in Q1 were race, age, gender, insurance type, and household income (all p < 0.05). More patients with NVAF switched/discontinued DOACs in Q1 vs. Q2-Q4, and more of them tended to remain untreated relative to those who switched/discontinued later in the year, suggesting a potential long-term impact of NMSD. Findings on factors associated with switching/discontinuation highlight potential socioeconomic discrepancies in treatment continuity.

Harnessing synthetic data to address fraud in cross-border payments

The sharing of data between financial institutions is widely recognised as a key component in the industry’s efforts to combat fraud. Broader access to multiple sources of financial data is also critical to the development of high-quality fraud detection mechanisms based on artificial intelligence (AI). Given the challenges relating to sharing real financial data across countries and institutions, the use of synthetic data has recently become critical to enabling the exploration of broader data sharing and supporting open collaboration in AI model development. To generate synthetic data that can substitute for real data, computer algorithms closely mimic the key statistical properties of genuine data, while strictly preserving the privacy and sovereignty of the source data. This paper presents the results of an ongoing exploration into the generation of high-utility synthetic datasets of cross-border payment transactions using transformer models and discusses its application to the development of AI-based fraud prevention solutions.

Enhancing fraud detection efficiency in mobile transactions through the integration of bidirectional 3d Quasi-Recurrent Neural network and blockchain technologies

Cases of financial fraud are increasing despite recent technical breakthroughs. It is tough to find authentic financial transaction data due to privacy issues and a lack of inter-organization synergy. However, for technologies based on data, such as machine learning to function accurately inside practical systems, they require actual data. This manuscript proposes a Fraud Detection (FD) Efficiency in Mobile Transactions through the Integration of Bidirectional 3D Quasi-Recurrent Neural Networks and Proof of Voting consensus Blockchain Technologies (Bi-3DQRNN-PoV-FD-MT). A dataset of Bitcoin transactions is employed in the suggested model. The banking sector’s Bitcoin transactions serve as the foundation for this dataset. Then the data balancing is performed using a Self-Adaptive Synthetic Over-Sampling Technique (SASOS). Then, the proposed framework utilizes the Bi-3DQRNN to determine if the information is fraudulent or not. Moreover, an intelligent Enhanced Artificial Gorilla Troops (EAGTO) Optimization algorithm is introduced to tune the weights of the model parameters. A blockchain-based proof of Voting (PoV) consensus algorithm is integrated with the proposed model for forecasting upcoming transactions. The developed scheme is instigated in Python and the performance provides 12.09%, 8.91%, and 6.92% higher accuracy compared with existing techniques like K-nearest neighbor- Distributed Blockchain Consortium (KNN-DBC), Decision tree-Ethereum blockchain-enabled smart contract (DT-EBSC) and Heterogeneous Graph Transformer Networks-Ethereum smart contract (HGTN-ESC).

Collateral requirements and corporate policy decisions

We challenge the theorized trade-off between risk management and investment due to collateral constraints. We compile a unique dataset of derivative transactions and collateral for U.S. public firms. Exploiting exogenous variation in cash-collateral, we observe significant effects on hedging but no impact on investment. Variations in PPE-collateral, instead, impact investment but show no association with hedging. Our findings suggest that a firm’s assets should not be seen as interchangeable; they rather play distinct roles in the collateralization process.

Applying the LightGBM model and SHAP interpretation to understand the urban land price index and its influencing factors in China

ABSTRACT Investigating the key factors affecting land prices is conducive to establishing targeted market-oriented reforms of the land market and promoting its healthy development. Traditional linear analysis cannot comprehensively compare the importance of feature variables and is unable to analyse the nonlinear influence process of feature variables. Therefore, based on the land transaction dataset from 2011 to 2020 in China, we used machine learning and the Shapley Additive exPlanations method to analyse the importance ranking of factors affecting land prices and analyse the nonlinear influence processes. The government fiscal, industrial development, and financial system indices are determined as the most important factors affecting residential, commercial, and industrial land price indices, respectively, and each exhibits different characteristics. The financial system index exerts the strongest impact on land price indices in large cities, whereas transportation and communication indices are the most dominant in small cities. Moreover, industrial development index is the most important influencing factor in industrial cities, whereas regional economic index exerts the greatest impact in small cities. By employing machine learning methods, we determined nonlinear effects that traditional linear models fail to obtain. Based on the results, we propose policy recommendations for market-oriented reforms on the land market, real-estate policies, and a stable monetary system.

Integrating a machine learning-driven fraud detection system based on a risk management framework

This article explores the application of machine learning techniques, specifically focusing on ensemble methods like Random Forests, for detecting fraudulent activities in digital financial transactions. Highlighting the evolution from traditional statistical approaches to modern machine learning models, it underscores the effectiveness of Random Forests in handling the inherent challenges of imbalanced datasets typical in fraud detection scenarios. Using a Kaggle dataset of credit card transactions, the study optimizes Random Forest parameters through rigorous parameter tuning, achieving significant improvements in model performance metrics such as Area Under the Curve (AUC). The findings underscore the critical role of machine learning in enhancing fraud detection capabilities, emphasizing the ongoing evolution and future potential of these methodologies in financial risk management.

Association Relationship Analysis in Finding Sales of Goods With Apriori Algorithm

Technology can be designed to help human life from all aspects ranging from agriculture, health science, industry and daily life. Toko Intan, a business engaged in the sale of basic and daily necessities. Every day, Toko Intan records every sales transaction in an archive stored in Microsoft Excel, containing data on goods sold every day. The purpose of this research is to find out what items are bought simultaneously by consumers to manage inventory, with the data mining method used in this research is the Association Rules method. Association Rules is one of the data mining techniques from the a priori algorithm which functions to find a combination pattern of an item. Tests carried out to process data in this study using the RapidMiner application, from tests carried out with the specified parameters, namely minimum support 30% and minimum confidence 65%, resulted in a lift ratio validation rule of 1.206. Personal Care Biscuits with 30.8% support and 90.9% confidence with a validation lift ratio of 1.206. Sales transaction data analysis can be applied well, and is able to generate a new association rule from the sales transaction dataset. With this research, it is hoped that it can provide information to the owner of the Intan store in providing the stock of goods needed by consumers and to find out the combination of item sets from the sales transaction dataset.

Robust Demand Estimation With Customer Choice-Based Models for Sales Transaction Data

We develop a novel statistical method to estimate customer choice among a firm’s portfolio of offerings when the firm cannot directly observe customers who choose not to purchase any product. This censored demand problem is prevalent in many industries such as hotels, airlines, and retail. Although several methods have been proposed to address this problem, they require some level of data aggregation across arrivals and/or choice sets, which results in information loss and potentially biased estimates. Therefore, they have limited applicability in an environment where the prices of a firm’s portfolio of offerings vary over time and sometimes even across different customers. Our proposed method combines several desirable properties, which makes it a better fit for realistic datasets where the available choice sets or attributes of the products in the choice sets change over time. We consider two additional types of information for identification of our model parameters: (1) additional mild assumptions on the customers’ utility function, and (2) external information about a firm’s market share. We then develop a robust estimation procedure that accounts for inaccuracies in either information type and let the data determine the best approach. Through Monte Carlo simulations, we show that our approach provides promising predictions of customer choice behavior when compared with other generally used methods and clearly outperforms those methods in scenarios where the product prices change frequently over time. Utilizing a real hotel transaction dataset provided by Oracle Labs, we further illustrate the improved estimation accuracy of our method compared to benchmark methods. Relative to existing approaches for estimating customer choice-based models, our proposed methodology better suits environments employing dynamic pricing and personalized offering practices, such as hospitality or online retailing.