Microsoft today announced the availability of R Server 9.0 on MSDN and Visual Studio Dev Essentials. This new release is built on top of the latest open source R engine, and comes with several new capabilities. In MRS 9.0, you can combine the algorithms delivered in this package with pre-existing parallel external memory algorithms such as the RevoScaleR package as well as open source tech like CRAN R packages to deliver the best predictive analytics.
MicrosoftML includes these algorithms:
- Fast linear learner, with support for L1 and L2 regularization.
- Fast boosted decision tree.
- Fast random forest.
- Logistic regression, with support for L1 and L2 regularization.
- GPU-accelerated Deep Neural Networks (DNNs) with convolutions.
- Binary classification using a One-Class Support Vector Machine.
Microsoft is releasing MicrosoftML on Windows and SQL Server today, with support for Linux and Hadoop to follow in the new year. MRS 9.0 now supports Spark 2.0, in addition to Spark 1.6, and also adds support for Ubuntu, complementing our support for SUSE and RedHat Linux. Microsoft is also today releasing a more powerful version of the R Client with these capabilities:
- New MRS packages, including MicrosoftML, OlapR, MRSDeploy and SQLrutils.
- Stay-current: We automatically check for updates and inform the user if a new release is available.
- Support for offline installation.
Read more about this release here.