1 IoT Fundamentals: Definitions, Architectures, Challenges, and Promises
25
• Trigger Alerts/Customized Workflows – Triggers a specific process or function in
response to an input pattern.
• Visualize Data – Data is sent to a Power BI (a business intelligence framework)
dashboard to allow real-time data visualization.
• Store Data – Utilizes Azure storage system to store the data; therefore, you
can perform batch analytics or train holistic machine learning models based on
historical data.
1.3.4.4 Azure Machine Learning
Azure Machine Learning is a cloud-based service supportive of open-source
technology and useful for large-scale training, deploying, automating, and managing
machine learning models. Azure Machine Learning enables the user to access
thousands of open-source Python packages that include machine learning components such as PyTorch, Scikit-learn, and TensorFlow. Microsoft also offers another
framework called Azure Machine Learning Studio, a drag-and-drop, a collaborative
area that allows you to create, test, and deploy machine learning solutions without
writing code. This workspace also provides preconfigured and pre-built algorithms
as well as data management modules that make experimenting with machine
learning modules quick and uncomplicated. Azure Machine Learning Service is
beneficial instead of Azure Machine Learning Studio, when greater control over
the details of the machine learning algorithms is needed or you need the flexibility
to utilize open-source machine learning libraries [16].
1.3.4.5 Azure Logic Apps
Azure Logic Apps is a cloud service used to arrange or automate tasks, workflow, or
business processes when data, applications, systems, or services must be integrated
across large enterprises. One of the main benefits of Azure Logic Apps is that it
makes the designing and implementation of scalable data integration, applications,
and other system solutions including business-to-business (B2B) communication
within the cloud or on premises (or both) easier and more straightforward. Below
you will find examples of workloads that is possible to automate using Azure Logic
Apps [16]:
• Event Processing – Events can be processed and routed across cloud services and
on-premises systems.
• Email Notification – Email notification can be automatically sent via Office 365
when an event occurs in an app, service, or system.
• File Transfer – Uploaded files can be transferred from FTP or SFTP servers to
Azure storage.
• Tweet Monitoring – Tweets can be reviewed by subject or analyzed based on
sentiment and alerts or tasks can be created if the additional inspection is
required.
Précédent

- 34/647

Suivant