Microsoft Azure Machine Learning

Chapter 11. Publishing a Model as a Web Service

So far, you have explored how to build different predictive models as experiments. Now, you might be wondering how that would be of use in real-life scenarios. One use of it is after you build and test your model, you can take a dataset and make a prediction straightaway inside the experiment, but in most of the cases, the result or the predictions of a model need to be used elsewhere and probably by some other people. Consider you have built a recommender model and it recommends items that a buyer might be interested in. So, these predictions require that the e-commerce site display the recommended product items to the prospective buyers. Consider one more scenario, where you have built a model ...