Microsoft Microsoft Certifications AI-900 Questions & Answers

• Question 151:

  You are building an AI-based app.

  You need to ensure that the app uses the principles for responsible AI.

  Which two principles should you follow? Each correct answer presents part of the solution.

  NOTE: Each correct selection is worth one point.

  A. Implement an Agile software development methodology

  B. Implement a process of AI model validation as part of the software review process

  C. Establish a risk governance committee that includes members of the legal team, members of the risk management team, and a privacy officer

  D. Prevent the disclosure of the use of AI-based algorithms for automated decision making

  Correct Answer: BC

  Reference: https://docs.microsoft.com/en-us/azure/cloud-adoption-framework/innovate/best-practices/trusted-ai https://docs.microsoft.com/en-us/learn/modules/responsible-ai-principles/3-implications-responsible-ai-practical

• Question 152:

  You have insurance claim reports that are stored as text.

  You need to extract key terms from the reports to generate summaries.

  Which type of AI workload should you use?

  A. natural language processing

  B. conversational AI

  C. anomaly detection

  D. computer vision

  Correct Answer: A

  Key phrase extraction is the concept of evaluating the text of a document, or documents, and then identifying the main talking points of the document(s). Key phase extraction is a part of Text Analytics. The Text Analytics service is a part of the Azure Cognitive Services offerings that can perform advanced natural language processing over raw text.

  https://docs.microsoft.com/en-us/learn/modules/analyze-text-with-text-analytics-service/2-get-started-azure https://docs.microsoft.com/en-us/azure/architecture/data-guide/technology-choices/natural-language-processing

• Question 153:

  You are building a knowledge base by using QnA Maker. Which file format can you use to populate the knowledge base?

  A. PPTX

  B. XML

  C. ZIP

  D. PDF

  Correct Answer: D

  D: Content types of documents you can add to a knowledge base:

  Content types include many standard structured documents such as PDF, DOC, and TXT.

  Note: The tool supports the following file formats for ingestion:

  .tsv: QnA contained in the format Question(tab)Answer.

  .txt, .docx, .pdf: QnA contained as regular FAQ content--that is, a sequence of questions and answers.

  Incorrect Answers:

  A: PPTX is the default presentation file format for new PowerPoint presentations.

  B: It is not possible to ingest xml file directly.

  Reference: https://docs.microsoft.com/en-us/azure/cognitive-services/qnamaker/concepts/data-sources-and-content

• Question 154:

  In which scenario should you use key phrase extraction?

  A. identifying whether reviews of a restaurant are positive or negative

  B. generating captions for a video based on the audio track

  C. identifying which documents provide information about the same topics

  D. translating a set of documents from English to German

  Correct Answer: C

• Question 155:

  In which two scenarios can you use a speech synthesis solution? Each correct answer presents a complete solution. NOTE: Each correct selection is worth one point.

  A. an automated voice that reads back a credit card number entered into a telephone by using a numeric keypad

  B. generating live captions for a news broadcast

  C. extracting key phrases from the audio recording of a meeting

  D. an AI character in a computer game that speaks audibly to a player

  Correct Answer: AD

  Azure Text to Speech is a Speech service feature that converts text to lifelike speech. Incorrect Answers:

  C: Extracting key phrases is not speech synthesis.

  Reference: https://azure.microsoft.com/en-in/services/cognitive-services/text-to-speech/

• Question 156:

  You use drones to identify where weeds grow between rows of crops to send an instruction for the removal of the weeds. This is an example of which type of computer vision?

  A. object detection

  B. optical character recognition (OCR)

  C. scene segmentation

  Correct Answer: A

  Object detection is similar to tagging, but the API returns the bounding box coordinates for each tag applied. For example, if an image contains a dog, cat and person, the Detect operation will list those objects together with their coordinates in the image.

  Incorrect Answers:

  B: Optical character recognition (OCR) allows you to extract printed or handwritten text from images and documents.

  C: Scene segmentation determines when a scene changes in video based on visual cues. A scene depicts a single event and it's composed by a series of consecutive shots, which are semantically related.

  Reference: https://docs.microsoft.com/en-us/ai-builder/object-detection-overview https://docs.microsoft.com/en-us/azure/cognitive-services/computer-vision/overview-ocr https://docs.microsoft.com/en-us/azure/azure-video-analyzer/video-analyzer-for-media-docs/video-indexer-overview

• Question 157:

  You are processing photos of runners in a race.

  You need to read the numbers on the runners’ shirts to identity the runners in the photos.

  Which type of computer vision should you use?

  A. facial recognition

  B. optical character recognition (OCR)

  C. semantic segmentation

  D. object detection

  Correct Answer: B

  Optical character recognition (OCR) allows you to extract printed or handwritten text from images and documents.

  Reference: https://docs.microsoft.com/en-us/azure/cognitive-services/computer-vision/overview-ocr

• Question 158:

  Which type of machine learning should you use to identify groups of people who have similar purchasing habits?

  A. classification

  B. regression

  C. clustering

  Correct Answer: C

  Clustering is a machine learning task that is used to group instances of data into clusters that contain similar characteristics. Clustering can also be used to identify relationships in a dataset

  Reference: https://docs.microsoft.com/en-us/dotnet/machine-learning/resources/tasks

• Question 159:

  What are two metrics that you can use to evaluate a regression model? Each correct answer presents a complete solution. NOTE: Each correct selection is worth one point.

  A. coefficient of determination (R2)

  B. F1 score

  C. root mean squared error (RMSE)

  D. area under curve (AUC)

  E. balanced accuracy

  Correct Answer: AC

  A: R-squared (R2), or Coefficient of determination represents the predictive power of the model as a value between -inf and 1.00. 1.00 means there is a perfect fit, and the fit can be arbitrarily poor so the scores can be negative.

  C: RMS-loss or Root Mean Squared Error (RMSE) (also called Root Mean Square Deviation, RMSD), measures the difference between values predicted by a model and the values observed from the environment that is being modeled. Incorrect Answers:

  B: F1 score also known as balanced F-score or F-measure is used to evaluate a classification model.

  D: aucROC or area under the curve (AUC) is used to evaluate a classification model.

  Reference: https://docs.microsoft.com/en-us/dotnet/machine-learning/resources/metrics

• Question 160:

  When you design an AI system to assess whether loans should be approved, the factors used to make the decision should be explainable. This is an example of which Microsoft guiding principle for responsible AI?

  A. transparency

  B. inclusiveness

  C. fairness

  D. privacy and security

  Correct Answer: A

  Achieving transparency helps the team to understand the data and algorithms used to train the model, what transformation logic was applied to the data, the final model generated, and its associated assets. This information offers insights about how the model was created, which allows it to be reproduced in a transparent way.

  Incorrect Answers:

  B: Inclusiveness mandates that AI should consider all human races and experiences, and inclusive design practices can help developers to understand and address potential barriers that could unintentionally exclude people. Where possible, speech-to-text, text-to-speech, and visual recognition technology should be used to empower people with hearing, visual, and other impairments.

  C: Fairness is a core ethical principle that all humans aim to understand and apply. This principle is even more important when AI systems are being developed. Key checks and balances need to make sure that the system's decisions don't discriminate or run a gender, race, sexual orientation, or religion bias toward a group or individual.

  D: A data holder is obligated to protect the data in an AI system, and privacy and security are an integral part of this system. Personal needs to be secured, and it should be accessed in a way that doesn't compromise an individual's privacy.

  Reference: https://docs.microsoft.com/en-us/azure/cloud-adoption-framework/innovate/best-practices/trusted-ai https://docs.microsoft.com/en-us/azure/cloud-adoption-framework/strategy/responsible-ai