39 / 54
W H I T E P A P E R
www.persistent.com
© 2017 Persistent Systems Ltd. All rights reserved.
10.3.2.4 Azure Machine Learning
Azure ML’s core feature is to make predictions on your data. This service provides studio which comes with lots of
drag-and-drop tool such as projects, web services, datasets, and notebooks you can use these tools build, test,
and deploy predictive analytics solutions.
Features
• Ability to load the dataset from sources such as delimited text files, SQL server, Azure storage services
and Hive tables etc.
• Azure ML has supports classification, regression, text analytics, clustering etc. (e.g. K-means, Linear
regression, SVM etc.) models.
• Ability to visualize the data using in the form of bar, histogram and bar plots etc.
• Ability to prepare/transform the data to make predications (such as feature selection, normalization, filter
and clean missing data)
• Allow to create web service of well-designed model and use it in some another application.
10.3.2.5 Azure Resource Manager
The Azure RM is specially designed to monitor, deploy and maintain the cloud resources. The resources might
be windows machine, Linux machine, storage services etc. Azure RM helps to deploy the solution as per your
required machine specification and resources for your use case (i.e. helps to deploy your solution).
Features
• Repeatedly deploy your solution.
• Can group the most relevant resources which helps organization to view the billing.
• Can set RBAC to services in your resource group.
10.3.2.6 Azure Function
The Azure function is small piece of code which runs in the cloud and can be used to carry out time based
processing of data (e.g. cleans up database table after every 15 minutes), develop REST API and trigger based
event processing, image processing, etc.
Features
• Developer can code in variety of languages.
• Developer can add his/her own dependencies (e.g. jar files).
• Easy integration with azure other services: these services can trigger your functions and can be used as
input or output for your functions.
• Function can be triggered on the occurrences of different event such as blob trigger, event hub trigger,
timer trigger, HTTP trigger etc.
10.3.2.7 Azure SQL Data Warehouse:
The SQL data warehouse is a scale-out database capable of processing massive volumes of data, both relational
and non-relational. It’s built on a massively parallel processing
(MPP)
Architecture, combining SQL Server
relational database technology with Azure cloud scale-out capabilities. The Azure SQL warehouse consist of
separate storage and compute nodes. It’s easy to deploy, seamlessly maintained, and fully fault tolerant because
of automatic back-ups. This service also provides security features such as
authentication, authorization and
data encryption.
39