Anti-Money Laundering: Anomaly Detection in Data

In recent years, anomaly detection has become an intensely studied topic in machine learning due to its wide range of applications. In particular, many financial institutions employ anomaly detection techniques in order to combat money laundering. 

In our research, we develop a new anomaly detection algorithm in a unsupervised setting that combines the metrics of distance and isolation, the Analytic Isolation and Distance-based Anomaly (AIDA) detection algorithm. AIDA is the first distance-based method that does not rely on the concept of nearest-neighbours, making it a parameter-free model. 

Moreover, when alerting of a possible illicit transaction, it is of paramount relevance to explain the reasoning behind such accusation. Hence, we also explore outlier explanation algorithms, and propose the Tempered Isolation-based eXplanation (TIX) algorithm, which finds the most relevant outlier features even in data sets with hundreds of dimensions. 

Researchers

Luis Souto Arias MSc and prof. Cornelis W. Oosterlee