Scholarlink

<p>Individuals can utilize credit cards for electronic transactions because they provide a convenient and safe alternative. The likelihood of credit card abuse has grown as their use has expanded.&#xa0;Credit card fraud causes considerable monetary losses for financial institutions and credit card users.&#xa0;This empirical investigation identifies fakes, examining the openness of open information, upper-class disparity information, modifications in fraud behaviour, and higher false rates. The intended work introduces a model that helps to validate credit card transactions to enhance consistency in the use of transactions and prevent distributed fraud detection in the financial system. Here, we used an ML-enabled technique in which feature selection is done using mutual classifier ANOVA, and a combination of under-sampling followed by Borderline SMOTE using a pipeline is used for data balancing. The data is split into an 80–20% train-test group, and Support Vector Machines (SVM), Random Forest (RF), and Decision Tree (DT) are used to build the model, along with cross-validation to detect anomalies. To intensify optimization, a brute force approach has been used, which provides the optimal solution. We fully interpret our model with a dataset of actual financial transactions. This pioneering research on artificial intelligence aims to enhance monitoring capabilities and efficiency in financial transactions, representing a major milestone in the persistent efforts to combat financial malpractice.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine Learning Framework Balancing Security and Efficiency in Digital Transactions

  • 期刊论文

  • 发表日期：

    2026年4月9日

查看全文
  • Jigyasha Arora,
  • Suyash Bhardwaj

摘要

Individuals can utilize credit cards for electronic transactions because they provide a convenient and safe alternative. The likelihood of credit card abuse has grown as their use has expanded. Credit card fraud causes considerable monetary losses for financial institutions and credit card users. This empirical investigation identifies fakes, examining the openness of open information, upper-class disparity information, modifications in fraud behaviour, and higher false rates. The intended work introduces a model that helps to validate credit card transactions to enhance consistency in the use of transactions and prevent distributed fraud detection in the financial system. Here, we used an ML-enabled technique in which feature selection is done using mutual classifier ANOVA, and a combination of under-sampling followed by Borderline SMOTE using a pipeline is used for data balancing. The data is split into an 80–20% train-test group, and Support Vector Machines (SVM), Random Forest (RF), and Decision Tree (DT) are used to build the model, along with cross-validation to detect anomalies. To intensify optimization, a brute force approach has been used, which provides the optimal solution. We fully interpret our model with a dataset of actual financial transactions. This pioneering research on artificial intelligence aims to enhance monitoring capabilities and efficiency in financial transactions, representing a major milestone in the persistent efforts to combat financial malpractice.