A Review on Machine Learning and Hybrid Ensemble Techniques for Money Laundering Detection in Financial Transactions

Abstract

Money laundering, a crime of severe financial dimension, tinges the tranquillity, transparency, and security of the global financial systems. Conventional anti-money laundering (AML) methodologies are to a large extent rule-based, making it considerably hard to detect ever-evolving, complex money-laundering patterns, thereby frequently making a final error. Now, with huge transactions on the digital banking sphere and the increased frequency in these transactions, clever data-driven mechanisms need to come into existence for identifying suspicious activity. This review walks through an extensive discussion on the detection of money laundering techniques through the use of rule-based systems, statistical methods, machine learning algorithms, or hybrid ensemble approaches. Machine learning approaches like logistic regression, decision tree, random forest, support vector machine, and XGBoost saw enhanced efficacy in making very good suggestions of patterns in the data of transactions. But these single models face problems with class imbalance, overfitting, and generalised data. Now, hybrid and ensemble learning models, which score better in single and multiple classifiers, are essential for ensuring accuracy, robustness, and reliability. Also, optimisation models such as genetic algorithms are given rings to improve the results of hyperparameter optimisation. The review also pays emphasis on good data preprocessing, feature extraction, and an essential ingredient in today's day—temporal data split and, in the end, performance metrics evaluation, which includes metrics like accuracy, precision, recall, F1 score, and ROC-AUC. This research has shown that a combination of the use of machine-learning models with ensemble machine-learning models and optimisation techniques is optimal for an extensible and efficient economic model for better day-front financial crime; it thus better targets financial crimes in real-world applications.

How to cite this article:
Mishra R, Rai A K. A Review on Machine Learning and Hybrid Ensemble Techniques for Money Laundering Detection in Financial Transactions. Int J Hum Comp Inter Data Min. 2026; 9(2): 24-32.