Fraud analytics with autoencoders
Fraud detection and prevention in financial companies such as banks, insurance companies, and credit unions is an important task. So far, we have seen how, and where, to use Deep Neural Networks (DNNs) and Convolutional Neural Network (CNNs).
Now it's time to use other unsupervised learning algorithm, such as autoencoders. In this section, we will be exploring a dataset of credit card transactions and trying to build an unsupervised machine-learning model that is able to tell whether a particular transaction is fraudulent or genuine.
More specifically, we will use autoencoders to pretrain a classification model and apply anomaly detection techniques to predict possible fraud. Before we start, we need to know the dataset.
Description of the dataset
For this example, we will be using the Credit Card Fraud Detection dataset from Kaggle. The dataset can be downloaded from https://www.kaggle.com/hunk3749/credit-card/data. Since I am using the dataset, it would...
