Monitoring adaptive models

Data scientists must regularly inspect the health of the out-of-the-box Pega Customer Decision Hub™ predictions and the adaptive models that drive them, and share their findings with the business team. The predictive performance and success rate of individual adaptive models provide information that can help business users and decisioning architects to refine business processes. Learn how to monitor the performance of predictions, adaptive models, and predictors.

Principiante

Inglés

Después de completar este módulo, podrá hacer lo siguiente:

Describe the lift metrics of a prediction

Name the key metrics of adaptive models visualized in the bubble chart.

Inspect individual active and inactive predictors.

Explain how predictors with similar predictive performance are grouped.

Examine the propensity distribution and the trend for the whole model.

Practique lo que ha aprendido en los siguientes Retos:

Monitoring predictions v5 Monitoring adaptive models v4

Disponible en la siguiente misión:

AI for 1:1 Customer Engagement v2

Temas

Regular monitoring of adaptive models

10 minutos

Monitoring predictions

10 minutos

Monitoring adaptive models

10 minutos

Cuestionario del módulo

10 minutos