Implementation of the Random Forest Algorithm for Financial Transaction Fraud Detection

Abstrak

Fraud detection in financial transactions is a major challenge for the banking and fintech industries, especially with the increasing volume of digital transactions. This study aims to implement the Random Forest algorithm in machine learning to detect suspicious financial transactions. The Random Forest algorithm was chosen due to its ability to handle complex data and produce accurate predictions. This research uses a financial transaction dataset consisting of various features such as transaction amount, location, payment method, and user activity patterns. The data undergoes preprocessing stages, including handling missing values, normalization, and oversampling techniques to address data imbalance. The Random Forest model is then developed and evaluated using accuracy, precision, recall, and F1-Score metrics to assess its fraud detection performance. The results show that the Random Forest model performs well in detecting fraudulent transactions with a high level of accuracy. Analysis of the confusion matrix also indicates that the model is able to reduce the number of false negatives, which is a crucial aspect in preventing losses due to illegal transactions. Additionally, a feature correlation heatmap is used to identify the most influential variables in fraud prediction. With the implementation of this Random Forest-based system, it is expected that the financial industry can enhance early detection of suspicious activities and strengthen security in digital transactions.

Identifier

• OpenAlex: https://openalex.org/W4415349533
• DOI: https://doi.org/10.59934/jaiea.v5i1.1635