Transaction Dataset Research Articles

Implementation of The Apriori Algorithm in Managing Stock Items at Drl.Rumahan Retail

Drl.Rumahan is a retail store that sells a variety of motorcycle lamp modifications. Drl.Rumahan is still struggling with determining stock levels and understanding customer purchases. Additionally, they are not utilizing transaction data as a valuable information source. Without leveraging this data, Drl.Rumahan will fall behind its business competitors and lose customers because the products they seek are unavailable. This situation will inevitably become a significant problem if it continues. This study aims to utilize sales transaction data as valuable information and identify customer purchasing patterns from the sales transaction data. The algorithm used is the Apriori algorithm to identify purchasing patterns from the transaction data set. The results of this study identified the three highest rules: if someone buys a pass beam switch, they will buy a shroud with a support value of 5.8% and a confidence value of 47.6%; if someone buys a shroud, they will buy a pass beam switch with a support value of 5.8% and a confidence value of 45.5%; and if someone buys a shroud, they will buy a relay with a support value of 5.2% and a confidence value of 40.9%. These results can inform business strategy decisions by increasing the inventory of products that form rules and serve as a guide for promotional product packages for products that have rules above the minimum support and minimum confidence.

Online payment fraud prediction with machine learning approach using naive bayes algorithm

The increase in e-commerce has provided easy access for the public, but it also opens up opportunities for fraud in online transactions. Payment fraud is also a problem that often arises in transactions through electronic media. This research aims to analyze payment fraud in e-commerce transactions. This research uses a machine learning approach using the Naive Bayes algorithm. This research uses online transaction datasets involving various attributes such as payment and shipping methods. The developed Naive Bayes model achieved an accuracy of 61.03% with K = 7. The evaluation shows a balance between precision (59.46%) and recall (62.05%), although this study is limited by data quality and basic assumptions of Naive Bayes. In future research, it is worth considering the use of additional features or more complex data processing to improve the performance of fraud detection in online transactions. This research provides important insights in the fight against financial crime in the context of electronic commerce.

Anomaly Detection in Blockchain Transactions: A Machine Learning Approach within the Open Metaverse

This study investigates the application of machine learning models for anomaly detection and fraud analysis in blockchain transactions within the Open Metaverse, amid the growing complexity of digital transactions in virtual spaces. Utilizing a dataset of 78,600 transactions that reflect a broad spectrum of user behaviors and transaction types, we evaluated the efficacy of several predictive models, including RandomForest, LinearRegression, SVR, DecisionTree, KNeighbors, GradientBoosting, AdaBoost, Bagging, XGB, and LightGBM, based on their Mean Cross-Validation Mean Squared Error (Mean CV MSE). Our analysis revealed that ensemble methods, particularly RandomForest and Bagging, demonstrated superior performance with Mean CV MSEs of -0.00445 and -0.00415, respectively, thereby highlighting their robustness in the complex transaction dataset. In contrast, LinearRegression and SVR were among the least effective, with Mean CV MSEs of -224.67 and -468.57, indicating a potential misalignment with the dataset's characteristics. This research underlines the importance of selecting appropriate machine learning strategies in the context of blockchain transactions within the Open Metaverse, showcasing the need for advanced, adaptable approaches. The findings contribute significantly to the financial technology field, particularly in enhancing security and integrity within virtual economic systems, and advocate for a nuanced approach to anomaly detection and fraud analysis in blockchain environments.

Implementasi Algoritma Apriori pada Market Basket Analysis terhadap Data Penjualan Produk Supermarket

Market Basket Analysis (MBA) is an analytical technique used to identify relationships between items in purchasing transactions. This notebook uses retail transaction datasets and the Apriori algorithm to discover hidden associations and patterns that retailers can leverage in optimizing marketing strategies, store layouts, and product recommendations. Through initial data processing, data exploration, and application of the Apriori algorithm, this analysis succeeded in identifying various significant associations between items that are frequently purchased together. The results provide valuable insights for retailers to develop targeted promotions and improve customer shopping experiences, while emphasizing the importance of selecting the right parameters to obtain accurate and relevant results.

Who Owns the Neighborhood? Ethnoracial Composition of Property Ownership and Neighborhood Trajectories in San Francisco

Property owners play pivotal roles in the trajectories of neighborhoods with discretion over upkeep, residential turnover, and affordability. Yet, little is known about how and why the racial composition of ownership changes over time relative to residents within a neighborhood and, in turn, how this relates to the neighborhood’s change and stability. With a self-constructed dataset of all residential transactions in San Francisco from 1990 to 2017, we consider how the ethnoracial composition of ownership differs from that of residents and how this difference relates to neighborhood change. We find that neighborhoods with more non-White residents have greater differences between the ethnoracial compositions of owners and residents, with the largest differences in neighborhoods with more Black residents. An increase in the divergence between these distributions is related to future increases in White and Asian residents and higher socioeconomic status residents and decreases in Black and Hispanic residents, illustrating that neighborhoods where owners are more ethnoracially distinct from the residents are more prone to neighborhood change and residential turnover. Our findings contribute to understandings of inequalities in property ownership and illuminate the role of ownership in neighborhood change in the contemporary city.

Holiday effects on tipping rates in full-service restaurants

Dining at restaurants on holidays involves a mostly different experience than dining on any other normal day, as it provides the opportunity for diners to get together with family and friends. Consequently, in this study, we investigate whether the prosocial behavior of restaurant diners deviates from the norm on holidays. Using a large dataset of credit card transactions at full-service restaurants, we confirm that this is indeed the case, albeit in a strikingly dichotomous fashion; religious and other family-oriented holidays are generally associated with significantly higher tipping rates than normal, whereas most other holidays are associated with significantly lower tipping rates than normal. Our findings suggest that family gatherings often meant to reinforce a sense of belonging and togetherness seem to strengthen prosociality, whereas social gatherings in restaurants on other festive occasions seem less likely to promote prosocial behavior.

An unsupervised machine learning approach to the spatial analysis of urban systems through neighbourhoods’ dynamics

Urban systems’ dynamics are the result of two intertwined processes that operate at different rhythms: their physical structure and underlying social processes. This paper suggests a novel approach to the spatial analysis of urban systems, using neighborhoods as a basic building block. Neighborhoods are usually the minimal homogeneous geographical unit in urban areas, both regarding their physical, and social characteristics, and the availability of governmental data. Using unsupervised machine learning algorithm, an extensive real-estate transaction dataset, and census data, a multi-scale analysis of neighborhoods’ dynamics in England and Wales is performed. The spatial and temporal dynamics of the resulting clusters of neighborhoods highlights the urban challenges faced by entire urban systems. The results suggest that processes triggered by urban inequalities may affect not only the social structure of cities (for example, through gentrification and displacement), but also the environmental sustainability of the whole urban system at a much larger scale: Suburbanization pressures seem to threaten rural areas at an unprecedented magnitude. The identification of the areas where this pressure is acute allows for the design of appropriate urban policy responses. The main message for the analysis of urban dynamics is that physical transformations and socio-political rearrangements that are intrinsically intertwined: Therefore, the joint management of the both aspects is the key to a sustainable future. The analysis also highlights the potential of machine learning algorithms for the benefit of urban science in general, and the study of urban dynamics in particular.

Avoiding Imprudent Commitments to Fast Dispatch: The Influence of Platform-Issued Quick-Seller Certificates on Postpurchase Behavior

ABSTRACT This study investigates the impact of platform-issued “quick-seller” certificates (indicating sellers with faster dispatch than others) on buyers’ postpurchase behavior. Drawing upon the theoretical lens of psychological contract violation (PCV), we hypothesize the effects of quick-seller certificates on order cancelations, as well as the subsequent impacts of cancelations associated with the quick-seller certificates on buyers’ future purchase frequency and search behavior. Using a comprehensive transaction dataset, the study reveals that quick-seller certificates increase the likelihood of cancelation of an order. Furthermore, we also find that an order cancelation from a certified quick seller results in decreased purchase frequency and a higher incidence of product comparisons through search engines.

Comparison of Classification Algorithm and Language Model in Accounting Financial Transaction Record: A Natural Language Processing Approach

The problem of financial recording not following the principles of accounting science has the potential to cause unnecessary problems. However, micro, small, and medium enterprises with their distinctive characteristics, though not all, still face many obstacles in writing financial reports. Even though there is already much financial software available, our study aims to investigate opportunities for implementing automation of accounting financial transaction records using the NLP approach, to interpret financial transactions based on text written on the transaction form into accounting journals (debits and credits). Experiments were carried out by comparing the performance of three classification algorithms, namely SVM, K-Nearest Neighbor, and Random Forest, with traditional (TF-IDF and BOW) and contextual (Word2Vec) Language Models. There are 200 financial transaction datasets consisting of ten classes. The data is divided into two parts, namely, the balance dataset and the imbalance dataset. The pair SVM and Word2Vec in the balanced dataset gave the highest accuracy (92.5%), precision (92.5%), recall/sensitivity (93.33%), and F1 score (92%). However, compared with the results of related semantic research (the average performance reaches 95%), the results obtained in this study are still lower. One point that may have a significant effect is the amount of data in the corpus, which is still lacking. Researchers suggest increasing the number of datasets and using a combination of other language models such as Glove, Bert etc. This study can also be used as a model for more complex financial transaction cases in future research.

Understanding Vulnerability to the Poverty Premium: An Analysis of Factors Influencing Use of High-Cost Credit Among Low-Income Individuals

Introduction & BackgroundAccess to affordable credit is essential for individuals living on low incomes to participate in society fully. However, due to their limited credit history and volatile incomes, many cannot access mainstream sources of credit, such as credit cards and personal loans. As a result, they often must rely on more expensive sources of credit such as payday loans, doorstep loans or rent-to-own loans. This leads to them paying more for credit, also referred to as the poverty premium. Incurring the poverty premium exacerbates the financial challenges faced by already vulnerable individuals and can lead to a cycle of financial distress. Identifying the behaviours and factors that lead to a need for high-cost credit can help in identifying individuals who are most vulnerable to incurring the poverty premium. Objectives & ApproachTo achieve this, we rely on anonymized Open Banking transaction data from 100,000 individuals provided by a UK-based social lender.Given the latent nature of vulnerability, we identify indicators of vulnerability to poverty premium which include frequency of overdraft use, previous debt problems, low financial resilience, and indebtedness. We use a copula-based approach to create an index of vulnerability to poverty premium. This is based on weighting the individual indicators using their Spearman rank correlation coefficient. We use a fixed effects model to identify the factors that contribute to this vulnerability, where the index of vulnerability is the dependent variable. Relevance to Digital FootprintsWe use a rich and granular dataset on individual financial transactions to address a key social issue. The findings from this study can inform policy and industry efforts to promote greater credit affordability for these vulnerable individuals. This is particularly important due to the renewed concerns regarding the increased use of high-cost credit by individuals living on low incomes due to COVID-19 and the increased cost of living. ResultsOur findings show that variables related to the financial profile of an individual are important driving factors of the vulnerability to poverty premium. These include the number of salary sources, frequency of salary receipt, benefit receipt and savings frequency. Other variables related to spending behaviour such as gambling, volatility in fixed expenses and high transaction counts all have positive relationships with this vulnerability. Conclusions & ImplicationsThis study is a first step towards examining the determinants of vulnerability to poverty premium by analyzing an Open Banking transaction data set. The innovative feature of this work is the creation of an index of vulnerability to poverty premiums based on various indicators of financial distress and high-cost credit use. Our findings on the relationships between the individual's financial profile and the vulnerability to poverty premium suggest that policymakers should consider targeting interventions for individuals with specific profiles to alleviate the need for high-cost credit. For example, programs to promote regular savings or financial education for gamblers and people with volatile spending behaviour can help assist these individuals manage their finances better.

Fraud Detection in Financial Transactions Using Credit Card: A Machine Learning Model

Fraud in any financial transactions causes severe loss to both customer as well as the seller. The loss is not confined to financial loss only but it also causes severe dent to the confidence of the customer especially related to shopping platform and the payment instrument used. Nowadays a substantial part of buying and selling of goods and services are taking place using various e commerce platforms. Many customer use credit card as payment instrument. During the payment process, users do exposes the credit card credentials to the payment platform. These payment platforms are vulnerable to fishing and hacking attackers and the payment instrument remains highly susceptible to fraudulent activities. There are two approaches to dealing with this problem. The first step is to identify the fraudulent activities and second step is to prevent any such attempt of fraud. This paper proposes a model based on machine learning algorithms to identify fraudulent attempts by analyzing credit card transactions data set and proposes methods of preventing such fraud activities. The paper also presents the accuracy of this AI model in identifying and preventing fraudulent and mischievous activities.

Attention of women's liberation and investor herding behavior

In recent decades, the women's liberation movement is garnering attention on a global scale. To investigate the impact of women's liberation attention on herding behavior among male and female investors in the stock market, we obtain a unique account-level data set of retail investor transactions in the Chinese stock market. Our results document that investors in Chinese cities with higher women's liberation attention demonstrate less herding behavior than other investors. Moreover, by facilitating investors' confidence and risk tolerance, women's liberation attention indirectly reduces herding behaviors. Our results remain solid in robustness tests, and additional tests confirm that women's liberation attention helps narrow the gender gap in herding behavior. Moreover, women's liberation attention is particularly significant in reducing investor herding behavior under volatile market conditions. Our paper contributes to the studies of gender equality in the stock market and provides empirical evidence by using account-level data.

Machine Learning Approaches for Credit Card Fraud Detection

Credit card fraud detection is a critical challenge in the financial sector, necessitating the adoption of advanced machine learning algorithms for timely and accurate identification of fraudulent transactions. In this project, we investigate the efficacy of four prominent machine learning algorithms - Support Vector Machines (SVM), K-Nearest Neighbors (KNN), Decision Trees (DT), and Random Forest (RF) in detecting credit card fraud. Through a comprehensive analysis, we evaluate the performance of these algorithms in terms of accuracy, precision, recall, and F1 score using real-world credit card transaction datasets. SVM, known for its ability to construct complex decision boundaries, excels in separating fraudulent and legitimate transactions. KNN leverages the proximity-based approach to identify similarities with known instances of fraud, while Decision Trees offer interpretable insights into fraudulent patterns. Random Forest combines the predictive power of multiple decision trees to produce robust and accurate predictions. Our findings shed light on the strengths and weaknesses of each algorithm, providing valuable insights for developing effective fraud detection systems in the financial industry

Spatial Constraints on Economic Interactions: A Complexity Approach to the Japanese Inter-Firm Trade Network

The trade distance is an important constraining factor underpinning the emergence of social and economic interactions of complex systems. However, agent-based studies supported by the granular analysis of distances are limited. Here, we present a complexity method that places the actual geographical locations of individual firms in Japan at the epicentre of our research. By combining methods derived from network science together with information theory measures, and by using a comprehensive dataset of Japanese inter-firm business transactions, we evaluate the effects of spatial features on the structural patterns of the economy. We find that the normalised probability distributions of the distances between interacting firms obey a power law like decay concomitant with the sizes of firms and regions. Furthermore, small firms would reach large distances to become customers of large firms, while trading between either only small firms or only large firms tends to be at smaller distances. Furthermore, a time evolution analysis suggests a reduction in the overall average trading distances in last 20 years. Lastly, our analysis concerning the trading dynamics among prefectures indicates that the preference to trade with neighbouring prefectures tends to be more pronounced at rural regions as opposed to the larger central conurbations.

Learning Transactional Behavioral Representations for Credit Card Fraud Detection.

Credit card fraud detection is a challenging task since fraudulent actions are hidden in massive legitimate behaviors. This work aims to learn a new representation for each transaction record based on the historical transactions of users in order to capture fraudulent patterns accurately and, thus, automatically detect a fraudulent transaction. We propose a novel model by improving long short-term memory with a time-aware gate that can capture the behavioral changes caused by consecutive transactions of users. A current-historical attention module is designed to build up connections between current and historical transactional behaviors, which enables the model to capture behavioral periodicity. An interaction module is designed to learn comprehensive and rational behavioral representations. To validate the effectiveness of the learned behavioral representations, experiments are conducted on a large real-world transaction dataset provided to us by a financial company in China, as well as a public dataset. Experimental results and the visualization of the learned representations illustrate that our method delivers a clear distinction between legitimate behaviors and fraudulent ones, and achieves better fraud detection performance compared with the state-of-the-art methods.

Fraud Feature Boosting Mechanism and Spiral Oversampling Balancing Technique for Credit Card Fraud Detection

With the flourishing of the credit card business and Internet technology, the risk of fraudulent credit card transactions is ever-increasing due to the complex information involved in the credit card business. Since the high redundancy of feature information and imbalance of class distribution in transaction data, the performance of the existing machine learning-based models for detecting credit card fraudulent transactions still needs to be improved. Therefore, it is crucial to build fraud detection models for effective feature engineering and sampling techniques. This article proposes a credit card fraud detection model incorporating a fraud feature-boosting mechanism with a spiral oversampling balancing technique (SOBT). Specifically, we present a compound grouping elimination strategy to exclude highly redundant and correlated features from the credit card transaction dataset and improve the data quality. Furthermore, we design a multifactor synchronous embedding mechanism, which combines the performance evaluation metrics of the embedding model for each feature and improves the decision-making ability of each feature for the target domain. Moreover, we propose an SOBT to balance the ratio of legitimate to fraudulent transactions, which improves the ability of the fraud detection model to distinguish legitimate from fraudulent transactions. Extensive experimental results based on two real-world datasets demonstrate that our methods can facilitate efficient credit card fraud detection and achieve better performance than state-of-the-art methods.

NUS: Noisy-Sample-Removed Undersampling Scheme for Imbalanced Classification and Application to Credit Card Fraud Detection

Since minority samples are substantially less common than majority samples, many industrial applications, such as credit card fraud detection (CCFD) and defective part identification, call for imbalanced classification. The performance of a classifier tends to suffer from the noisy samples in majority or minority classes. This work proposes a new undersampling scheme, called a clustering-based noisy-sample-removed undersampling scheme (NUS) for imbalanced classification. The majority class samples are first clustered. The distance of the majority class sample from the cluster center that is furthest away is used as the radius to build a hypersphere, with each cluster’s center assumed to be a spherical center. We determine the Euclidean distance between the center of a cluster and each minority sample to find whether they are in the hypersphere or not. Afterward, we exclude noisy samples from the minority class. The noisy samples of majority classes are removed by using the same procedure. Second, we propose an NUS, which combines noisy sample removal with undersampling techniques. Finally, to prove the effectiveness of NUS, we integrate NUS with the basic classifiers random forest (RF), decision tree (DT), and logistics regression (LR). We conduct their comparison with seven undersampling, oversampling, and noisy-sample-removed methods. This work performs experiments on 13 public and three real transaction datasets related to e-commerce. The results show that NUS plays a positive role in promoting existing classifiers’ performance.

A machine learning‐based approach for wait‐time estimation in healthcare facilities with multi‐stage queues

AbstractDigital technologies have been contributing to providing quality health care to patients. One aspect of this is providing accurate wait times for patients waiting to be serviced at healthcare facilities. This is naturally a complex problem as there is a multitude of factors that can impact the wait time. However, the problem becomes even more complex if the patient's journey requires visiting multiple stations in the hospital; such as having vital signs taken, doing an ultrasound, and seeing a specialist. The authors aim to provide an accurate method for estimating the wait time by utilising a real dataset of transactions collected from a major hospital over a year. The work employs feature engineering and compares several machine learning‐based algorithms to predict patients' waiting times for single‐stage and multi‐stage services. The Random Forest algorithm achieved the lowest root mean squared error (RMSE) value of 6.69 min among all machine learning algorithms. The results were also compared against a formula‐based system used in the industry, and the proposed model outperformed the existing model, showing improvements of 25.1% in RMSE and 18.9% in MAE metrics. These findings indicate a significant improvement in the accuracy of predicting waiting times compared to existing techniques.

Mining Top-k High On-shelf Utility Itemsets Using Novel Threshold Raising Strategies

High utility itemsets (HUIs) mining is an emerging area of data mining which discovers sets of items generating a high profit from transactional datasets. In recent years, several algorithms have been proposed for this task. However, most of them do not consider the on-shelf time period of items and negative utility of items. High on-shelf utility itemset (HOUIs) mining is more difficult than traditional HUIs mining because it deals with on-shelf-based time period and negative utility of items. Moreover, most algorithms need minimum utility threshold ( min_util ) to find rules. However, specifying the appropriate min_util threshold is a difficult problem for users. A smaller min_util threshold may generate too many rules and a higher one may generate a few rules, which can degrade performance. To address these issues, a novel top-k HOUIs mining algorithm named TKOS ( T op- K high O n- S helf utility itemsets miner) is proposed which considers on-shelf time period and negative utility. TKOS presents a novel branch and bound-based strategy to raise the internal min_util threshold efficiently. It also presents two pruning strategies to speed up the mining process. In order to reduce the dataset scanning cost, we utilize transaction merging and dataset projection techniques. Extensive experiments have been conducted on real and synthetic datasets having various characteristics. Experimental results show that the proposed algorithm outperforms the state-of-the-art algorithms. The proposed algorithm is up to 42 times faster and uses up-to 19 times less memory compared to the state-of-the-art KOSHU. Moreover, the proposed algorithm has excellent scalability in terms of time periods and the number of transactions.

Multiscale Feature Fusion and Graph Convolutional Network for Detecting Ethereum Phishing Scams

With the emergence of blockchain technology, the cryptocurrency market has experienced significant growth in recent years, simultaneously fostering environments conducive to cybercrimes such as phishing scams. Phishing scams on blockchain platforms like Ethereum have become a grave economic threat. Consequently, there is a pressing demand for effective detection mechanisms for these phishing activities to establish a secure financial transaction environment. However, existing methods typically utilize only the most recent transaction record when constructing features, resulting in the loss of vast amounts of transaction data and failing to adequately reflect the characteristics of nodes. Addressing this need, this study introduces a multiscale feature fusion approach integrated with a graph convolutional network model to detect phishing scams on Ethereum. A node basic feature set comprising 12 features is initially designed based on the Ethereum transaction dataset in the basic feature module. Subsequently, in the edge embedding representation module, all transaction times and amounts between two nodes are sorted, and a gate recurrent unit (GRU) neural network is employed to capture the temporal features within this transaction sequence, generating a fixed-length edge embedding representation from variable-length input. In the time trading feature module, attention weights are allocated to all embedding representations surrounding a node, aggregating the edge embedding representations and structural relationships into the node. Finally, combining basic and time trading features of the node, graph convolutional networks (GCNs), SAGEConv, and graph attention networks (GATs) are utilized to classify phishing nodes. The performance of these three graph convolution-based deep learning models is validated on a real Ethereum phishing scam dataset, demonstrating commendable efficiency. Among these, SAGEConv achieves an F1-score of 0.958, an AUC-ROC value of 0.956, and an AUC-PR value of 0.949, outperforming existing methods and baseline models.