What's new in Machine Learning Server
Applies to: Machine Learning Server (Find R Server 9.x What's New)
Machine Learning Server 9.2.1 expands upon Microsoft R Server 9.1 with new Python libraries for integrating machine learning and data science into analytical solutions in the enterprise. Language-specific development is available when you add Python or R support (or both) during setup.
Read our release announcement for Machine Learning Server 9.2.1.
For features in R Client, see What's New for Microsoft R Client.
In Machine Learning Server, Python support is through libraries that can be used in script executing locally, or remotely in either Spark over Hadoop Distributed File System (HDFS) or in a SQL Server compute context.
Python libraries include revoscalepy, microsoftml, and azureml-model-management-sdk. Modules are built on Anaconda 4.2 over Python 3.5. You can run any 3.5-compatible library on a Python interpreter included in Machine Learning Server.
Together, the libraries provide full-spectrum data mining: data transformation and manipulation, analysis and visualization, model management. Machine learning algorithms, as well as pre-trained models provided by Microsoft, are now in Python.
revoscalepy is a library provided by Microsoft to support distributed computing, local compute context, remote compute context for SQL Server and Spark, and high-performance algorithms for Python, similar to RevoScaleR.
Originally introduced in SQL Server 2017, this library is extended in Machine Learning Server to support a remote compute context for Spark 2.0-2.1 over the Hadoop Distributed File System (HDFS), with Python functions that run jobs in parallel across multiple nodes. This tutorial gets you started.
Pre-trained models for image classification and sentiment detection articulated in Python.
microsoftml machine learning algorithms and data mining.
Remote execution is not available for Python scripts. For information about to do this in R, see Remote execution in R.
R function libraries are built on Microsoft R Open (MRO), Microsoft's distribution of open source R 3.4.1.
R realtime model scoring is now also supported on Linux.
The last several releases of R Server added substantial capability for R developers. To review recent additions to R functionality, see feature announcements for previous versions.
Role-based access control (RBAC) has been extended with a new explicit Reader role.
Register your compute nodes with your web nodes in a centralized and simplified way in the Administration Utility.
Visit Feature announcements in R Server version 9.1 and earlier, for descriptions of features added in recent past releases.