What's New in SAS Model Manager 14.2

You can import SAS analytic store models that are created using SAS Factory Miner or by using the HPFOREST and HPSVM procedures in Base SAS. You can use a SAS analytic store model (SASAST) file to import an analytic store model into SAS Model Manager. However, SAS analytic store models that are created by using SAS Enterprise Miner and that are in SAS package file (SPK) format are not supported in this release. For more information, see Import a Model from a SAS Analytic Store File.

You can publish, score, monitor performance, or run reports for SAS analytic store models that are located within a project. SAS analytic store models can be published only to Hadoop and Teradata. Retraining of SAS analytic store models is not supported. For more information, see the following:

Users who are configured to use single sign-on web authentication by Kerberos can publish models to Hadoop and Teradata. For more information, see Configure Users Authenticated by Kerberos for Publishing Models in SAS Model Manager: Administrator’s Guide.

When you are creating a scoring test for a model, you can process the scoring test data in a High-Performance Analytics production environment. For more information, see Create a Scoring Test.

SAS Model Manager portfolio macros enable you to score SAS Factory Miner models that are within a SAS Model Manager portfolio. For more information, see %MM_ScorePortfolioModels Macro in SAS Model Manager: Macro Reference.

SAS Model Manager portfolio macros enable you to publish SAS Factory Miner models that are within a SAS Model Manager portfolio to a database or Hadoop. For more information, see %MM_PublishPortfolioModelsDB in SAS Model Manager: Macro Reference and %MM_PublishPortfolioModelsHadoop in SAS Model Manager: Macro Reference.