Examples and Tutorials related to fraud detection using machine learning and deep learning.
A beginner tutorial consisting of 3 parts show how to build a classifier with TensorFlow to detect fraudulent credit card transactions. We see how to develop a logisitic regression from scratch and extend it to a simple deep feed-forward network using TensorFlow's low-level api. At the end the tutorial gives an outlook on Keras.
This tutorial shows how to use using PyTorch to build a LSTM network with embeddings which can learn user profiles and detect anomalies in user activities. It's an example for using deep learning to detect anomalies. This method is a unsupervised learning tasks. And dosn't rely on prelabeled data.
These kind of method could be for example used to detect data theft from application logs or detect online shopping fraud.