Fraud Detection workshop with Red Hat OpenShift AI

Introduction

Welcome! In this workshop, you learn how to use data science, artificial intelligence (AI), and machine learning (ML) in an OpenShift development workflow.

You complete the following tasks in Red Hat OpenShift AI without installing any software on your computer:

  • Explore a pre-trained fraud detection model by using a Jupyter notebook.

  • Deploy the model by using OpenShift AI model serving.

  • Refine and train the model by using automated pipelines.

  • Learn how to train the model by using distributed computing frameworks.

About the example fraud detection model

The example fraud detection model monitors credit card transactions for potential fraudulent activity. It analyzes the following credit card transaction details:

  • The geographical distance from an earlier credit card transaction.

  • The price of the current transaction, compared to the median price of all the transactions.

  • Whether the user completed the transaction by using the hardware chip in the credit card, by entering a PIN number, or by making an online purchase.

Based on this data, the model outputs the likelihood of the transaction being fraudulent.

Before you begin

You must have access to an OpenShift cluster which has Red Hat OpenShift AI installed.

Due to resource constraints, this main (the latest) edition of the tutorial is not supported in the Red Hat Developer Sandbox.

If your cluster uses self-signed certificates, before you begin the workshop, your OpenShift AI administrator must add self-signed certificates for OpenShift AI as described in Working with certificates (Self-Managed).

If you’re ready, start the workshop.