ticlazau/german-credit

Notebooks and synthetic data for AI OpenScale demo with German Credit data

Notebooks and synthetic data for AI OpenScale demo with German Credit data

The notebooks and data in this repo will allow you to explore the capabilities of IBM AI OpenScale. There are two main notebooks to work with:

• The lab notebook will allow you to train, save and deploy the German Credit Risk model, and configure AI OpenScale to monitor that deployment.
• The end-to-end notebook does all of the above, and additionally uses Apache Spark with Watson Studio to configure continuous learning for your model.

Finally, you can use the credit feed notebook to regularly send data to your models, allowing for a continuous feed of predictions for OpenScale to monitor.

Right-click on the notebook you wish to run and select Save As... to download the file. To import the notebook into Watson Studio:

• Open Watson Studio
• Click Create a project and select a Standard project
• Give your project a name, select an existing object storage instance or create a new one, and click Create
• Once your project has been created, click on the Assets tab, click on Add to project at the top of the screen, and select NOTEBOOK
• On the New Notebook screen, select From file, choose the notebook file you downloaded, select the Default Spark Python... option, and click Create notebook

You may now continue following the instructions contained in the notebook.