This course provides an overview of the Risk Modeling nodes in SAS Visual Machine Learning. It prepares members of your organization to develop scorecards. The course includes hands-on demonstrations and teaches you key concepts, terminology, and base functionality that are integral to visual risk modeling.

Naucz się use SAS Visual Machine Learning to:

• Add data to the pipeline.
• Insert an Interactive Grouping node into the pipeline and specify grouping options.
• Add a Scorecard node to generate a scorecard.
• Control the selection and optimization option of the logistic regression model.
• Manually adjust bins and scorecard cutoff values.
• Insert a Reject Inference node into the pipeline.

Kto powinien uczestniczyć

Modelers, credit analysts, and others who are interested in risk modeling

Dostępne formaty	Długość
e-szkolenie:	8 godziny/180 dzień licencja

Before attending this course, you should have completed the SAS Viya: Uczenie maszynowe z wykorzystaniem Model Studio course. Other prerequisites are knowledge of and experience with logistic regression, which you can gain by completing the Statystyka 1: wprowadzenie do analizy wariancji, regresji i regresji logistycznej and Prognozowanie z użyciem Model Studio w SAS Viya courses.

To szkolenie wykorzystuje oprogramowanie SAS Viya  

This course addresses Risk Modeling Add-on for SAS Visual Machine Learning in SAS Viya.

Introduction to Risk Modeling Add-on

• Define credit risk.
• Define scorecards.
• Understand the risk modeling workflow.

Interactive Grouping Node

• Understand the purpose of grouping.
• Identify measures used to determine the overall predictive power of a characteristic.
• Add an Interactive Grouping node to the pipeline.
• Manage bins interactively.

Scorecard Node

• Understand basic logistic regression selection methods.
• Add a Scorecard node to the pipeline.
• Manage scorecard points and ranges interactively.

Reject Inference Node

• Define the reason for using a Reject Inference node.
• Add a Reject Inference node to the pipeline.