Scholarlink

In this data science study, we intended to apply machine learning algorithms in R to build a classifier that can detect credit card fraud. For analysis, this work used the card transactions dataset with both fraudulent and non-fraudulent transactions. Four machine learning algorithms were used to execute Logistic Regression (LR), Decision Trees (DTs), Artificial Neural Networks (ANNs), and Gradient Boosting Machines (GBMs) classifiers on the data. Each model was assessed using accuracy, precision, recall, F1-score, Receiver Operating Characteristic (ROC) curve, and Area Under the Curve (AUC). The best-performing model was selected and the findings were discussed while studying the findings. With an accuracy rate of 95.2%, the precision score was also optimum at 91%, recall was optimal at 90%, and F1-score was at 91%, which put the GBM as the ideal method for detecting fraud for this given dataset.

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

Strategic Fraud Detection in Credit Card Transactions Through Machine Learning Insights

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Abdulkareem Dawah Abbas,
  • Hussein Muhi Hariz,
  • Umar Farooq Khattak,
  • Riyadh Rashid Hameed,
  • Mustafa Mohammed Mahmood,
  • Dulfikar Jawad Hashim

摘要

In this data science study, we intended to apply machine learning algorithms in R to build a classifier that can detect credit card fraud. For analysis, this work used the card transactions dataset with both fraudulent and non-fraudulent transactions. Four machine learning algorithms were used to execute Logistic Regression (LR), Decision Trees (DTs), Artificial Neural Networks (ANNs), and Gradient Boosting Machines (GBMs) classifiers on the data. Each model was assessed using accuracy, precision, recall, F1-score, Receiver Operating Characteristic (ROC) curve, and Area Under the Curve (AUC). The best-performing model was selected and the findings were discussed while studying the findings. With an accuracy rate of 95.2%, the precision score was also optimum at 91%, recall was optimal at 90%, and F1-score was at 91%, which put the GBM as the ideal method for detecting fraud for this given dataset.