Application of machine learning algorithms on HADOOP platform for efficient analysis and classification of fraudulent transactions in the banking industry

The article uses the latest research work in the field of machine learning, big data, algorithms in this area and financial activities as research material. A number of analytical methods were applied in the process: the monographic method allowed to dive deeply into the topic, the evaluative method allowed effectively analyze the data, and the reflective method allowed to critically evaluate the findings. The use of machine learning in financial systems is a revolutionary approach to fraud detection and prevention. This method not only solves complex problems associated with unbalanced data and pattern variability, but also allows information to be processed with incredible speed and accuracy. As a result, the use of such technologies is not just improving, but revolutionizing the way financial institutions protect themselves from fraudulent attacks, ensuring the highest level of security for operations and customers.