Skip to content

Latest commit

 

History

History
30 lines (20 loc) · 1.36 KB

workshop_fraud.adoc

File metadata and controls

30 lines (20 loc) · 1.36 KB

Credit Card Fraud Detection Demo

Currently, this use case is a demo rather than a workshop. You can quickly deploy the demo implementation and demo or play around with it.

There’s no step-by-step instructions on how to build this, yet. If you are willing to contribute with the workshop content this will be greatly appreciated. Please get in touch.

Preparation

This demo uses MapBox to show the distribution of fraud occurrences on an interactive map. To enable this a MapBox token is required so that Cloudera DataViz can connect to the MapBox service. If you don’t yet have a token, click on the MapBox link above, sign up for free and create a token for yourself.

To deploy the demo please SSH to your cluster host and run the following command:

Note
The command below will undo everything done in the cluster in previous workshops.
MAPBOX_TOKEN=<your_mapbox_token> /tmp/resources/reset-to-lab.sh fraud 99

Demo points

After the demo is deployed you can browse the different components of the use implementation in the following UIs:

  • NiFi - Generate fake transaction data, processes it and forward it to Kafka and Kudu

  • Kafka/SMM - Browse data in Kafka topics

  • Hue - Browse data in Kudu tables

  • SSB/Flink - Check real-time streaming jobs

  • CDSR/DataViz - Browse the Fraud dashboard and datasets