SAS Model Manager is
a flexible model repository and model management environment for predictive
and analytical models. A centralized repository and procedural templates
make it easy to manage models and metadata across organizational areas
and throughout a model's life cycle. Accountability metrics and validation
of analytical steps, from creation to deployment in real time or batch
scoring systems, continue until a model is retired. Storing the models
in a secure, centralized repository enables you to easily manage life
cycle milestones, such as development, test, production, and retirement.
The following figure illustrates the model management process that
you use in SAS Model Manager:
The goal of a modeling
project is to identify a champion model that a scoring application
uses to predict an outcome. SAS Model Manager provides tools to evaluate
candidate models, declare a project champion model, and inform your
scoring officer that a predictive or analytical model is ready for
validation or production. You can perform scoring tests for champion
and challenger model assessment as well as publish and share the model
life cycle and performance data over established reporting channels.
You can also run comparative performance benchmarks for the models
in your production environment.