Summary
Phew! We did a lot of things in this chapter. First of all, we created, with few clicks, an Azure Databricks cluster. We exploited that large amount of computation power to train a logistic regression classifier to distinguish between spam and ham SMS messages. Azure Databricks helps us a lot, simplifying the process of model training, testing, and building.
After this, we entered into the realm of Azure Machine Learning, learning its key features with which we can simplify the development of a machine learning project. We then integrated Azure Machine Learning into Azure Databricks, improving our model by testing it with different regularization parameters. After persisting all the metrics and trained models inside Azure Machine Learning for future use, we created an image and deployed our best model—a web service.
Azure Databricks and Azure Machine Learning provide data scientists with great tools for training anything from simple to complex machine learning models. Computation...
