Bitcoin Abnormal Transaction Detection Based on Machine Learning

Feldman, Elena V.; Ruchay, Alexey N.; Matveeva, Veronica K.; Samsonova, Valeria D.

This paper is devoted to the development of a reliable abnormal bitcoin transaction detection that may be involved in money laundering and illegal traffic of goods and services. The article proposed an algorithm of abnormal bitcoin transaction detection based on machine learning. For training and evaluation of the model, the Elliptic dataset is used, consisting of 46564 Bitcoin transactions: 4545 of “illegal” and 42019 of “legal”. The proposed algorithm for detecting abnormal bitcoin transactions is based on various machine learning algorithms with the selection of hyperparameters. To evaluate the proposed algorithm, we used the metric of accuracy, precision, recall, F1 score and index of balanced accuracy. Using the resampling algorithm in conditions of class imbalance, it was possible to increase the reliability of the classification of abnormal bitcoin transactions in comparison to the best known result on the Elliptic dataset.