The Application Of Data Mining In Anti-money Laundering System

With the rapid development of global economic integration process, rampant money laundering in the financial sector has a great impact on the normal operation of the economy. Several related anti-money laundering laws and regulations are formulated all around the world, and the financial institutions are required to build the anti-money laundering systems. Under this circumstance, this thesis analyzed the related anti-money laundering model depending on the realistic financial system, and applied an anti-money laundering risk rating model in the system.The paper mainly focuses on the topic of data mining techniques applied in the anti-money laundering system, and proposes the anti-money laundering risk rating model. In this model, the paper adopt the induction classification technique from the data mining field and design an ensemble classifier that is based on the decision tree induction and rule-based induction methods. The anti-money laundering risk rating model uses the ensemble classifier to predict the customers’ money laundering risk level result using their attribute information stored in the financial institutions, and the customers’ risk level has the value of high, middle, and low. Then, we optimize the proposed ensemble classifier with the pruning and adaptive boosting techniques.The paper evaluates the proposed anti-money laundering risk rating model through detailed experiment and uses the customer validation set to evaluate the model,compares base classifiers with the ensemble model using reasonable evaluation metrics.The experimental results show that the ensemble classifier used in the risk rating model is more accurate. The paper analyzes the structure and relationship of anti-money laundering system and analyzes the function of SMBC-AML anti-money laundering system. The paper introduces the customer identification of large amount, suspect and key suspect classes according to the national anti-money laundering law,and applies the proposed model into the bank anti-money laundering system.The practical experience shows that the anti-money laundering risk rating model has a good applied prospect.