Retraining a New web service using PowerShell cmdlets

## Why retrain: defining the problem As part of the machine learning training process, a model is trained using a set of data. Models you create using Machine Learning are typically not static. As new data becomes available or when the consumer of the API has their own data the model needs to be retrained. ## End-to-end workflow The process involves the following components: A Training Experiment and a Predictive Experiment published as a web service. To enable retraining of a trained model, the Training Experiment must be published as a web service with the output of a trained model. This enables API access to the model for retraining. To retrain the model: * Update training experiment to allow for retraining * Deploy the retraining web service * Use the Batch Execution Service code to retrain the model For a walkthrough of the preceding steps, see [Retrain Machine Learning models programmatically](https://azure.microsoft.com/en-us/documentation/articles/machine-learning-retrain-models-programmatically/). Applying the retrained model: * Sign in to your Azure Resource Manager account * Get the Web service definition object * Export the Web Service Definition object as JSON * In the JSON, update the reference to the ilearner blob that contains the trained model * Import the JSON into a Web Service Definition object * Update the web service with new Web Service Definition object The following diagram shows the full process of retraining a model.  For a walkthrough of the preceding steps, see [Retrain an existing Predictive Web service](https://azure.microsoft.com/en-us/documentation/articles/machine-learning-retrain-existing-resource-manager-based-web-service/). The [AML Retraining Models Using APIs](https://www.youtube.com/watch?v=wwjglA8xllg) video shows you how to retrain Machine Learning models created in Azure Machine Learning using the Retraining APIs and PowerShell.