About Managing Models :: SAS® Model Manager 14.2: User's Guide

Using SAS Model Manager, you can store models, and organize them within projects or folders, validate candidate models, assess candidate models for champion model selection, and publish and monitor champion models in a production environment, and retrain models. All model development and model maintenance personnel, including data modelers, validation testers, scoring officers, and analysts, can use SAS Model Manager.

Here are some of the services SAS Model Manager provides:

Use a single interface to access all of your business modeling projects. All models are stored in a central, secure model repository. Models can also be accessed in one place using the model inventory list in the Inventory category.
Track the progress of your project’s version by creating processes, definitions, and tests. You create custom processes, definitions, and tests to meet your business requirements and to match your business processes.
Use data tables that are registered in the SAS Metadata Repository.
Import SAS Enterprise Miner models, SAS/STAT linear models, SAS/ETS COUNTREG and SEVERITY models, models that you develop using SAS code, PMML models, and R models. You can also import a generic model and the model’s files in to a folder. You can create custom model templates for SAS code models so that SAS Model Manager knows exactly what files and metadata are associated with a model.
You can schedule and run scoring tests, performance monitoring, and retraining to validate models.
Run several reports to compare and assess candidate models. You can also write your own SAS reporting programs to run and assess candidate models. The aggregated reporting facility enables you to combine multiple reports into a single report. Dashboard reports enable you to monitor the state of projects using performance monitoring reports and can be viewed in a web browser.
Publish models to the SAS Metadata Repository or a SAS channel. You can also publish the champion model and challenger models to a database for scoring. The SAS Scoring Accelerator is used by SAS Model Manager to publish models to a database or Hadoop.

Data tables are an integral part of the modeling process. You can use project input and output prototype tables, as well as scoring input and output prototype tables to define variables. Data tables are used for scoring, testing, and performance monitoring. Performance data can be created from your operational data, provided that it has the required structure (for example, the data contains a target variable).

You can also create multiple projects in a portfolio. Additional versions can then be created for all projects within the portfolio. Champion models for all projects within the portfolio can be monitored for performance, and published to the SAS Metadata Repository. SAS Factory Miner models can also be registered to the SAS Model Manager model repository. The SAS Factory Miner projects are managed as portfolios in SAS Model Manager. The project segments and models are available within a portfolio and can be managed from the Portfolios category within SAS Model Manager.

Any user who is registered in SAS Management Console can be assigned to a SAS Model Manager group, and can then work in SAS Model Manager. For more information, see Configuring Users, Groups, and Roles in SAS Model Manager: Administrator’s Guide.