خديجه طبايي

One of the main challenges in the banking system is fraud and money laundering. Fraud and money laundering cause dirty money from illegitimate activities to enter the economic cycle of the country. Fraud and money laundering have destructive effects on the banking system and cause significant losses to the banking system. One of the modern solutions to deal with money laundering and bank fraud is the use of blockchain technology, but in most studies, this method has been used to conduct secure transactions and cannot detect fraud. Tools such as machine learning and data mining are needed to examine the content of transactions to detect fraud. In this research, a hybrid approach has been presented to deal with and detect money laundering transactions. In the proposed method, bank customers are first authenticated and validated with blockchain, and in the second step, with several data mining methods, the pattern of fraud in bank transactions can be detected. In the stage of knowledge discovery, three data-mining methods are used for majority voting, in such a way that a transaction is a fraud when two or three transaction methods are recognized as a fraud. The eva‎luations show that the average index of accuracy, sensitivity and precision of the proposed method for detecting fraud and money laundering is 96.68%, 95.84% and 95.87% respectively. Experiments in Weka and MATLAB showed that the accuracy of the proposed method is more accurate than the three classifiers of multilayer neural network, support vector machine and random forest in detecting money laundering. The proposed method is more accurate in detecting bank fraud than the combined methods of genetics, nearest neighbor and the combined method of particle swarm optimization algorithm and gravity search. Using blockchain increases the security of the proposed method in authentication and has high scalability

اعتبارسنجي مشتريان , مشتريان بانكي , تقلب و پولشويي , بلاك‌چين , داده‌كاوي , راي گيري اكثريت

Credit Assessment of customers , Banking Customers , Fraud And Money Laundring , Blockchain , Data mining , Majority voting

Khadijeh Tabaei

Rahim KhaniZad