Fraud and Anomaly Detection with Artificial Neural Networks using Python3 and Tensorflow.

Over the last few years, there has been a increasing trend in demand for the application of anomaly detection models within the field of data science -- especially when it comes to the detection of fraudulent vs non-fraudulent actions. Within the following dataset, we will explore the use of a number of different predictive models, each with varying complexity. As with every good data science project, we will first examine the dataset, preprocess our data, explore the contents, train a number of models, and finally review and evaluate the results. For the purposes of this tutorial, the dataset we will be using consists of credit card transactions by European cardholders. You can access the dataset as well as the code from my GitHub.