I have designed a SOM for fraud detection in a Credit Card Customer Dataset. SOM being unsupervised learning approach with no labels try to categorize customers together and areas marked in white in SOM visualisation are the Fraud Customers. Using reverse transformation on Min max scaler I am able to get the Customer ID of fraud customers.
mpenam/Deep_Learning-Self_Organizing_Map_for_fraud_detection
Self Organizing Map (Python) I have designed a SOM for fraud detection in a Credit Card Customer Dataset. SOM being unsupervised learning approach with no labels try to categorize customers together and areas marked in white in SOM visualisation are the Fraud Customers. Using reverse transformation on Min max scaler I am able to get the Customer ID of fraud customers.
Jupyter Notebook