Question 1
Question
What is edge computing?
A. An architecture that processes data as close to its source as possible
B. A new name for computing
C. A type of computing that leaves network teams on edge
D. Computing that teams can only attempt when standing on the edge of something
Answer
A. An architecture that processes data as close to its source as possible
Explanation
Edge computing architecture processes client data at a network’s edge, which is as close to the data’s source as possible. Edge computing eliminates the distance between users and applications, which alleviates many bandwidth, latency and throughput issues that distances can cause. This architecture combines data center and cloud architectures and alleviates their inefficiencies — such as the distance between users and applications — for a more modern, efficient technology.
Question 2
Question
True or false: Edge computing and the network edge are the same thing.
A. True
B. False
Answer
B. False
Explanation
Edge computing and the network edge are not the same. While edge computing needs a network edge to function, the network edge does not require edge computing. A network edge is one or more perimeters that separate a network into parts: the parts the organization owns and the parts third parties operate. Edge computing processes and stores data at the network edge, but the network edge exists with or without edge computing.
Question 3
Question
What’s the difference between edge computing and fog computing?
A. Experts use them interchangeably
B. The architectures place intelligence and compute power in separate places
C. Fog computing encompasses a large web of connected devices and data locations, and edge computing processes data and compute solely at the edge
D. All of the above
Answer
D. All of the above
Explanation
The answer to this question gets a little foggy. Network professionals have varied responses to this question, hence why all answers could be correct. Some use fog computing and edge computing interchangeably, others believe the two differ slightly and some others believe they differ greatly.
Like edge computing, fog computing processes data and compute closer to the data source. Yet, instead of processing data specifically at the edge, fog computing processes data between the source and the cloud. Both fog and edge computing bring cloud-like capabilities closer to data processing, but the processing methods differ.
Question 4
Question
What is edge computing’s role in cloud computing?
A. They are the same
B. Edge computing is the cloud’s silver lining — or the edge of the cloud
C. Edge computing could act as an alternative to cloud computing
D. They are unrelated
Answer
C. Edge computing could act as an alternative to cloud computing
Explanation
Edge computing takes after cloud computing in many ways, but they are not the same. Cloud computing delivers hosted services over networks — in particular, the internet. Edge computing won’t completely replace cloud computing, but edge computing is less likely to cause delays from data moving across the network — like cloud computing could — because it processes data at the edge. The two also differ in terms of security, as edge computing doesn’t store data at distributed sites, so hackers have less access to vulnerable data.
Question 5
Question
What technological advancement contributed to edge computing’s popularity?
A. IoT
B. The cloud
C. 5G
D. 11ax, or Wi-Fi 6
Answer
A. IoT
Explanation
IoT — the internet of things — played a crucial role in enterprise interest in edge computing, as IoT and edge computing together force organizations to reconsider their traditional network architectures and how they process IoT device data. The IoT market also continues to grow, as does the amount of IoT devices worldwide. With so many devices that require data processing, network access and the necessary throughput, organizations can turn to edge computing to address these issues. Edge computing and IoT together can help organizations cut down on WAN transport costs and alleviate quality of service problems.
Question 6
Question
What issue is common in many edge computing challenges?
A. Bandwidth
B. Latency
C. Network security
D. Network traffic
Answer
D. Network traffic
Explanation
Some common edge computing challenges include distributed computing, security and data accumulation — and traffic plays a role in all of them. Inconsistent traffic patterns can complicate security and compute models, which may force network teams to reconsider their network configurations. Other challenges include network bandwidth, latency and data backup.
Question 7
Question
True or false: Edge computing offers quick response times and support for large data amounts.
A. True
B. False
Answer
A. True
Explanation
Two main edge computing benefits include faster response times and increased support for large and growing amounts of data. Other benefits include centralized management, with lights-out — or remote — capabilities and cloud-based infrastructure, which can help network teams deliver local services more easily.
Question 8
Question
What are the benefits of edge computing security?
A. Edge computing security only benefits IoT
B. Edge computing only secures data about edges, including mountain edges and table edges
C. Edge computing security can respond in real time and host behavioral threat analytics
D. Edge computing cannot be secure
Answer
C. Edge computing security can respond in real time and host behavioral threat analytics
Explanation
While edge computing security has challenges, the architecture stores data at the edge and prevents bad actors from accessing vulnerable data at individual sites. And edge computing can secure IoT, which helps teams secure an organization’s network as a whole. Edge computing can also host secure web access gateways — which allow devices to connect to the organization’s overall network — and behavioral threat analytics, which provide an additional tier of management and monitoring.
Question 9
Question
How can software-defined networking (SDN) work with edge computing?
A. Edge computing only works in a software-defined network
B. SDN can streamline how edge computing processes data
C. They don’t work well together
D. They are the same
Answer
B. SDN can streamline how edge computing processes data
Explanation
With SDN and edge computing, SDN can help decide whether to process data and tasks at the edge or in the cloud. SDN can help make effective data processing decisions for individual networks and their desired outcomes and even alleviate bottlenecking issues. Over time, the two will likely interact and intertwine further.
Question 10
Question
Why should anyone care about edge computing?
A. It can alleviate latency issues
B. It can ease network congestion
C. It can bolster bandwidth for IoT devices
D. All of the above
Answer
D. All of the above
Explanation
If your network deals with any of the issues above, then edge computing matters for you. Edge computing’s ability to process data closer to data sources can reduce latency, as data no longer travels to distant data centers or clouds, which takes up unnecessary time. Edge computing can alleviate congestion because it performs its operations locally, so raw data movement no longer overwhelms networks. And bandwidth — especially for IoT devices — maintains a better overall performance, even in areas with unreliable connectivity.
Question 11
Question
Analytics professionals and consultants have identified two up-front requirements for predictive analytics initiatives to be successful. These two requirements are:
A. Clearly defined business objectives and investment in the right professional talent
B. Carefully selected predictive modeling tools and IT teams willing to learn new skills
C. Mathematically inclined employees and the latest predictive analytics software
Answer
A. Clearly defined business objectives and investment in the right professional talent
Explanation
Without clear business reasons for instituting an analytics program, predictive analytics can be a wasted effort. In addition, predictive modeling typically must be done by data scientists and other analytics professionals with specialized skills.
Question 12
Question
True or false? Creating predictive models is an iterative process that requires ongoing development and refinement.
A. True
B. False
Answer
A. True
Explanation
Developing successful predictive models is an intricate undertaking that requires gathering input from business managers up front and then testing and refining the models to ensure that they work properly.
Question 13
Question
One recommended strategy for implementing predictive analytics without breaking the bank is:
A. Hire a team of skilled analytics professionals to manage your predictive modeling software.
B. Build and test prototype predictive models based on pertinent business questions.
C. Choose a tool that can handle all potential needs so you don’t have to make future purchases.
Answer
B. Build and test prototype predictive models based on pertinent business questions.
Explanation
Consultants offer several suggestions for implementing predictive analytics on a tight budget in this article, including the strategy of building and testing a prototype predictive modeling system before making a full investment.
Question 14
Question
Combining traditional data warehousing and big data analytics in a hybrid approach to analytics enables companies to do ad hoc queries and manage the flood of unstructured data while also:
A. Supporting the needs of data scientists building complex predictive models
B. Replacing conventional business intelligence functionality with new reporting capabilities
C. Creating algorithms and models from historical, structured information
Answer
A. Supporting the needs of data scientists building complex predictive models
Explanation
Bringing together traditional data warehousing and business intelligence approaches and big data analytics in a hybrid scenario allows for conventional BI operations as well as new capabilities for analyzing large and varied data sets. This approach can also facilitate the work of data scientists.
Question 15
Question
True or false? The goal of predictive analytics is to study the past but ultimately to generate predictions about the future.
A. True
B. False
Answer
A. True
Explanation
Predictive analytics gives data analysts a way to connect past events with potential future business outcomes.
Question 16
Question
According to Neil Ward-Dutton, research director at U.K.-based MWD Advisors, event processing and predictive analytics complement each other. All of the following are examples of applications that mix event processing and predictive analytics, EXCEPT:
A. A retailer analyzing inventory losses to identify possible causes
B. A hospital using patient-monitoring data to assess the likelihood that individual patients may contract blood poisoning
C. A manufacturer predicting factory machine failures based on sensor data and preordering replacement parts
D. A financial services company tracking website logs and correlating online activity to possible security threats
Answer
A. A retailer analyzing inventory losses to identify possible causes
Question 17
Question
True or false? It’s easy for most companies to find employees with advanced analytics skills.
A. True
B. False
Answer
B. False
Explanation
While there are certainly many skilled data analysts out there, the increasing adoption of predictive analytics, big data analytics and other forms of advanced analytics by businesses is making the hunt for people with the required skills more competitive.
Question 18
Question
Which two cloud providers joined forces to create Gluon?
A. IBM and Microsoft
B. AWS and Google
C. Google and IBM
D. AWS and Microsoft
Answer
D. AWS and Microsoft
Explanation
In October 2017, AWS and Microsoft teamed up to create Gluon, an open source deep learning library. Its interface and automation capabilities aim to make it easier for developers to build machine learning models.
Question 19
Question
True or false: Machine learning and deep learning are the same thing.
A. True
B. False
Answer
B. False
Explanation
While both AI technologies are similar, machine learning and deep learning have different applications. Machine learning enables software to learn and predict outcomes, such as how much wood a construction company might need to order in the spring, without additional programing. Deep learning, also known as deep neural networking, takes it a step further and looks deeper into data for trends and relationships. A deep learning service, for instance, could make recommendations on which movie you’d like to watch based on your viewing habits.
Question 20
Question
Which suite of services from Microsoft offers APIs to embed image and language processing capabilities into an app?
A. Azure Smart Cloud
B. Microsoft Cognitive Services
C. Azure Artificial Intelligence
D. Microsoft AI Services
Answer
B. Microsoft Cognitive Services
Explanation
Microsoft Cognitive Services offers a number of machine learning technologies that enable a developer to incorporate capabilities such as image processing and speech-to-text into an application. The suite includes software development kits and over 20 APIs.
Question 21
Question
Which AWS managed service aims to help enterprises more easily and quickly integrate machine learning-based models into applications?
A. Amazon SageMaker
B. Amazon Fargate
C. Amazon MagicMan
D. Amazon Sorcerer
Answer
A. Amazon SageMaker
Explanation
At AWS:reInvent 2017, the company introduced Amazon SageMaker. The machine learning and AI service primarily targets developers and aids with the creation, training and management of machine learning models. It comes with 10 of the most common machine learning algorithms built in.
Question 22
Question
Which product is at the center of IBM’s machine learning and AI portfolio?
A. IBM Moriarty
B. IBM Holmes
C. IBM Sherlock
D. IBM Watson
Answer
D. IBM Watson
Explanation
IBM Watson is a supercomputer that is at the core of the vendor’s cognitive services. With IBM Watson and its APIs, enterprises can create chatbots, analyze data, turn speech into text, identify emotions through text and more.
Question 23
Question
Which AWS service infuses machine learning into cloud security?
A. Amazon Stacie
B. Amazon Macie
C. Amazon Incognito
D. Amazon Cognito
Answer
B. Amazon Macie
Explanation
Amazon Macie is an automation tool that uses machine learning to discover, classify and protect data in the cloud provider’s Simple Storage Service. It monitors data and sends an alert in the event of suspicious activity. Amazon Macie can also automatically take action against certain threats.
Question 24
Question
True or false: Azure Machine Learning Studio doesn’t require programming to build predictive analysis models.
A. True
B. False
Answer
A. True
Explanation
Azure Machine Learning Studio has an interactive, visual interface with drag-and-drop abilities to construct, test and deploy predictive analysis models — no programming required. Developers can choose from a library of machine learning algorithms and access a gallery that shows examples of real-world applications. Once developers build a model, they can publish it as a web service.
Question 25
Question
Which Google-developed service competes most directly with Gluon?
A. TensorFlow
B. Kaggle
C. Dataflow
D. Tenor
Answer
A. TensorFlow
Explanation
TensorFlow is an open source machine learning software library used to build computational graphs. Developed internally at Google, TensorFlow offers numerous models for object detection, voice recognition, translation and more.
Question 26
Question
Which of the following best describes predictive modeling?
A. A process marketers use to evaluate how factors influence future behavior
B. A process of building models that predict the future for businesses
C. A predictive analytics process that creates a statistical model of future behavior
Answer
C. A predictive analytics process that creates a statistical model of future behavior
Explanation
While predictive modeling is often used by marketing agencies and departments, it also has many other potential uses for predicting future behavior.
