MicrosoftML (MML) is a new machine learning package for Microsoft R Server. Microsoft R Server brings you the ability to do parallel and chunked data processing that addresses the restrictions of in-memory open source R. MML adds Microsoft’s battle-tested algorithms and data transforms that are used by product teams across Microsoft. This brings new machine learning functionality with increased speed, performance and scale, especially for handling a large corpus of text data and high-dimensional categorical data.
MML includes the ability to:
- Create text classification models for problems such as sentiment analysis and support ticket classification.
- Train deep neural nets in order to solve complex problems such as retail image classification and handwriting analysis.
- Work with high-dimensional data for scenarios like online advertising click-through prediction.
- Solve many other common machine learning tasks such as churn prediction, loan risk analysis, and demand forecasting using state-or-the-art, fast and accurate algorithms.
- Integrate with SQL Server to easily pull data from SQL and train in R and use your trained model inside SQL.
In this webinar, we will describe the machine learning capabilities in Microsoft R Server and show several demos of how you can begin using it today!