Machine Learning

Version: 10.0

The Sitecore Cortex™ Processing Engine (Processing Engine) provides an integration point for your machine learning platform of choice.

Microsoft Machine Learning Server integration

The Microsoft Machine Learning Server integration allows you to send your machine learning script and the location of projected data to a Machine Learning Server web endpoint. Integrating with Machine Learning Server includes the follow high-level tasks:

  • Implement a custom model wrapper:

    • Inject IScriptSessionFactory into the model.

    • In the context of the TrainAsync() and EvaluateAsync() methods, use IScriptSessionFactory  to send your R script and the location of your training data to a Machine Learning Server endpoint.

  • To use the training logic within your custom model, schedule a training worker task that references your model. The training worker executes the TrainAsync() method.

    Note

    Out-of-the box integration with Microsoft Machine Learning Server has been deprecated and will be removed in a future release. You can still supply data that has been processed by Cortex to a machine learning engine of your choice.

Machine learning and Experience Optimization

Machine learning in the context of Experience Optimization includes:

  • A custom PersonalizationInteractionModel (implements IModel ) that projects interaction data. The TrainAsync() and EvaluateAsync() methods are not implemented.

  • A custom BuildPersonalizationRulesWorker that fits a decision tree to projected data and deploys the results to the xDB Reporting database. Decision tree logic is contained within an injected class.

Do you have some feedback for us?

If you have suggestions for improving this article,