2020/05/29 by Joana Lorenz, Maria Inês Silva, Lorenz, Joana +7 · 4 citations
Computer Science · Social Sciences · #Blockchain Technology Applications and Security #Crime, Illicit Activities, and Governance #FOS: Computer and information sciences #Imbalanced Data Classification Techniques #Machine Learning (cs.LG) #Machine Learning (stat.ML)
paper · pdf · doi:10.48550/arxiv.2005.14635
openalex publication_date 2020/05/29 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Every year, criminals launder billions of dollars acquired from serious\nfelonies (e.g., terrorism, drug smuggling, or human trafficking) harming\ncountless people and economies. Cryptocurrencies, in particular, have developed\nas a haven for money laundering activity. Machine Learning can be used to\ndetect these illicit patterns. However, labels are so scarce that traditional\nsupervised algorithms are inapplicable. Here, we address money laundering\ndetection assuming minimal access to labels. First, we show that existing\nstate-of-the-art solutions using unsupervised anomaly detection methods are\ninadequate to detect the illicit patterns in a real Bitcoin transaction\ndataset. Then, we show that our proposed active learning solution is capable of\nmatching the performance of a fully supervised baseline by using just 5 % of\nthe labels. This solution mimics a typical real-life situation in which a\nlimited number of labels can be acquired through manual annotation by experts.\n