An Unsupervised Learning approach to identify frauds in Credit Card applications using Self Organised Maps
There are two Jupyter Notebooks having two different approaches to detection of potential frauds in Credit Card applications.
First approach is using a simple Self Organised Map technique using MiniSOM model.
Second approach is a combination of first approach and adding a supervised Artificial Neural Network (ANN) to build a more robust model for potential frauds in Credit Card applications.