APMG International ARTIFICIAL-INTELLIGENCE-FOUNDATION Online Practice Questions and Exam Preparation

APMG International ARTIFICIAL-INTELLIGENCE-FOUNDATION Online Questions & Answers

• Question 11:

  Ensemble learning methods do what with the hypothesis space?

  A. Select a combination of hypothesis to combine their predictions
  B. Use stochastic gradient descent to optimise a network.
  C. Extract ergodic solutions.
  D. Test multiple hypotheses simultaneously.
  A. Select a combination of hypothesis to combine their predictions

  ---

  Explanation/Reference:

  https://link.springer.com/referenceworkentry/10.1007/978-0-387-73003- 5_293#:~:text=Definition,and%20combine%20them%20to%20use. It works by selecting different subsets of the data, or different combinations of the hypothesis, and

  combining the results of each prediction in order to create a single, more accurate result. This is useful in situations where different hypothesis may be accurate in different parts of the data, or where a single hypothesis may not be accurate in

  all cases. Ensemble learning is used in a variety of applications, from computer vision to natural language processing.

  References:

  [1] BCS Foundation Certificate In Artificial Intelligence Study Guide, BCS

  [2] Apmg-international.com, "What is Ensemble Learning?", APMG International, https://apmg- international.com/en/about-apmg/blog/what-is-ensemble-learning/

  [3] Exin.com, "Ensemble Learning", EXIN, https://www.exin.com/en-us/learn/ensemble-learning

• Question 12:

  What does TRL stand for?

  A. Technical Robotic Level.
  B. Transform Reinforced Learning
  C. Technology Readiness Level.
  D. Transport Ready Level.
  C. Technology Readiness Level.

  ---

  Explanation/Reference:

  Technology Readiness Level (TRL) Technology Readiness Levels (TRL) are a method of estimating the technology maturity of Critical Technology Elements (CTE) of a program during the acquisition process.

  https://acqnotes.com/acqnote/tasks/technology-readiness- level#:~:text=Technology%20Development,Technology%20Readiness%20Level%20(TRL),program%20during%20the%20acquisition %20process.

  TRL stands for Technology Readiness Level and is a measure of how close a technology is to being ready for use in a real-world environment. TRL is used to assess the progress of research and development of a technology, ranging from

  basic research (TRL 1) to fully operational (TRL 9). TRL is used to help determine the level of completion of a technology and its potential success in a real-world environment.

  References:

  [1] https://www.bcs.org/upload/pdf/foundation-certificate-ai-syllabus-v1.pdf

  [2] https://www.apmg-international.com/en/qualifications-and-certifications/bc-foundation-certificate-in-artificial-intelligence/

  [3] https://www.exin.com/en/certifications/bc-foundation-certificate-in-artificial-intelligence/

  [4] https://www.acq.osd.mil/rd/nii/trl.html

• Question 13:

  With a large dataset, limited computational resources or frequent new data to learn from, we can adopt what type of machine learning?

  A. Batch learning.
  B. Big Data learning.
  C. Patchwork learning.
  D. Online learning.
  D. Online learning.

  ---

  Explanation/Reference:

  Online learning is a type of machine learning that can be used when a large dataset is limited in computational resources or if the data is frequently changing. It allows the system to learn from new data as it is being presented, rather than having to re-train the entire dataset each time new data is added. This makes it more efficient and effective than batch learning, as it only needs to process the new data and not the entire dataset. Online learning is often used in applications such as fraud detection, where new data is constantly being added and needs to be analyzed quickly. For more information, please refer to the BCS Foundation Certificate In Artificial Intelligence Study Guide (https://www.bcs.org/upload/pdf/bcs-foundation-certificate-in-artificial-intelligence-study-guide.pdf) or the EXIN Artificial Intelligence Foundation Certification (https://www.exin.com/en/exams/artificial-intelligence-foundation).

• Question 14:

  Reflex and Model-based Reflex are two types of what?

  A. Robot
  B. Artificial intelligent agents.
  C. Algorithms.
  D. Compilers.
  B. Artificial intelligent agents.

  ---

  Explanation/Reference:

  Reflex and Model-based Reflex are two types of Artificial Intelligent Agents. Artificial Intelligent Agents are computer systems designed to act and think in a manner similar to humans,incorporating elements of problem solving, decision-

  making, communication, and learning. Reflex agents are reactive agents which act based on the current environment and conditions, while Model-based Reflex agents use a model of the environment to make decisions.

  References: BCS Foundation Certificate In Artificial Intelligence Study Guide, https://bcs.org/ai/certificate/ and APMG International, https://www.apmg-international.com/qualifications/artificial-intelligence- foundation-certificate.

• Question 15:

  What technique can be adopted when a weak learners hypothesis accuracy is only slightly better than 50%?

  A. Over-fitting
  B. Activation.
  C. Iteration.
  D. Boosting.
  D. Boosting.

  ---

  Explanation/Reference:

  Weak Learner: Colloquially, a model that performs slightly better than a naive model.

  More formally, the notion has been generalized to multi-class classification and has a different meaning beyond better than 50 percent accuracy. For binary classification, it is well known that the exact requirement for weak learners is to be

  better than random guess. [...] Notice that requiring base learners to be better than random guess is too weak for multi-class problems, yet requiring better than 50% accuracy is too stringent.

  -Page 46, Ensemble Methods, 2012.

  It is based on formal computational learning theory that proposes a class of learning methods that possess weakly learnability, meaning that they perform better than random guessing. Weak learnability is proposed as a simplification of the

  more desirable strong learnability, where a learnable achieved arbitrary good classification accuracy. A weaker model of learnability, called weak learnability, drops the requirement that the learner be able to achieve arbitrarily high accuracy; a

  weak learning algorithm needs only output an hypothesis that performs slightly better (by an inverse polynomial) than random guessing.

  -The Strength of Weak Learnability, 1990.

  It is a useful concept as it is often used to describe the capabilities of contributing members of ensemble learning algorithms. For example, sometimes members of a bootstrap aggregation are referred to as weak learners as opposed to

  strong, at least in the colloquial meaning of the term.

  More specifically, weak learners are the basis for the boosting class of ensemble learning algorithms.

  The term boosting refers to a family of algorithms that are able to convert weak learners to strong learners.

  https://machinelearningmastery.com/strong-learners-vs-weak-learners-for-ensemble- learning/

  The best technique to adopt when a weak learner's hypothesis accuracy is only slightly better than 50% is boosting. Boosting is an ensemble learning technique that combines multiple weak learners (i.e., models with a low accuracy) to create

  a more powerful model. Boosting works by iteratively learning a series of weak learners, each of which is slightly better than random guessing. The output of each weak learner is then combined to form a more accurate model. Boosting is a

  powerful technique that has been proven to improve the accuracy of a wide range of machine learning tasks. For more information, please see the BCS Foundation Certificate In Artificial Intelligence Study Guide or the resources listed above.

• Question 16:

  Narrow or weak Al can be useful to robots.

  Which of the following is an example of narrow Al?

  A. Conscious simul-ation.
  B. Artificial General Al.
  C. Conscious integration.
  D. NLP - Natural Language Processing.
  D. NLP - Natural Language Processing.

  ---

  Explanation/Reference:

  NLP - Natural Language Processing is an example of narrow AI. It is a type of AI system that is able to understand, interpret, and generate natural language. NLP has become increasingly popular over the past few years, as it has been used

  to create applications such as chatbots, virtual assistants, and search engines. NLP systems are able to learn language and the context in which it is used, and they are able to understand the nuances of language and its different meanings.

  References:

  BCS Foundation Certificate In Artificial Intelligence Study Guide, https://bcs.org/certifications/foundation- certificates/artificial-intelligence/

• Question 17:

  The EU's Ethical Guidelines use what to demonstrate trustworthy Al?

  A. A quality assurance plan.
  B. UN's sustainability goals.
  C. Customer feedback.
  D. A human-centric value system.
  D. A human-centric value system.

  ---

  Explanation/Reference:

  The European Union's Ethical Guidelines for Trustworthy AI use a human- centric value system to demonstrate that Artificial Intelligence (AI) is trustworthy. This value system is based on human rights, autonomy, safety, privacy, transparency,

  accountability and fairness. The guidelines also state that AI should be designed, developed and used in a manner that respects these values.

  References:

  https://ec.europa.eu/digital-single-market/en/news/ethical-guidelines-trustworthy-ai

  BCS Foundation Certificate In Artificial Intelligence Study Guide (2019), A.I and Ethics, Chapter 5.

• Question 18:

  If Al undertakes routine and monotonous tasks and takes these away from humans, what will humans do?

  A. Higher value work.
  B. Leisure activities
  C. Change jobs.
  D. Sabotage the Al.
  A. Higher value work.

  ---

  Explanation/Reference:

  Al is designed to take on routine and monotonous tasks, freeing up humans to take on more complex, higher value work. This can include tasks such as research, problem-solving, and decision-making. This shift in work roles is expected to

  increase productivity and efficiency, allowing humans to focus on more creative and innovative tasks. For example, robots can be used to automate mundane manufacturing processes, freeing up human workers to take on jobs that require

  more creative thinking and problem- solving.

  References:

  [1] https://www.bcs.org/upload/pdf/foundation-certificate-ai-syllabus-v1.pdf

  [2] https://www.apmg-international.com/en/qualifications-and-certifications/bc-foundation-certificate-in-artificial-intelligence/

  [3] https://www.exin.com/en/certifications/bc-foundation-certificate-in-artificial-intelligence/

• Question 19:

  How could machine learning make a robot autonomous?

  A. Use OCR, optical character recognition, to read documents
  B. Use NLP (Natural Language Processing) to listen
  C. Use actuators to modify its environment
  D. Learn from sensor data and plan to carry out a task.
  D. Learn from sensor data and plan to carry out a task.

  ---

  Explanation/Reference:

  Machine learning can be used to make robots autonomous by allowing them to learn from sensor data and plan how to carry out a task. This involves using algorithms to analyze data from sensors and use this data to make decisions and

  take actions. By using machine learning, robots can learn from their environment and become more autonomous.

  References:

  [1] BCS Foundation Certificate In Artificial Intelligence Study Guide, "Robotics", p.98.

  [2] APMG-International.com, "Foundations of Artificial Intelligence"

  [3] EXIN.com, "Foundations of Artificial Intelligence"

• Question 20:

  Healthcare can benefit from Al, and in particular Machine Learning, an example of which is?

  A. Autonomous wheelchairs.
  B. Automated blood sampling.
  C. Autonomous vehicles.
  D. Diagnostic image analysis
  D. Diagnostic image analysis

  ---

  Explanation/Reference:

  Healthcare can benefit from AI, and in particular Machine Learning, in a number of ways. One example is diagnostic image analysis, which can help to automatically identify and classify abnormalities in medical images such as X-rays, CT scans, and MRI scans. Machine Learning algorithms can be used to detect patterns in the data which can be used to accurately diagnose diseases and illnesses.

  References:

  [1] https://www.bcs.org/upload/pdf/foundation-certificate-ai-syllabus-v1.pdf

  [2] https://www.apmg-international.com/en/qualifications-and-certifications/bc-foundation-certificate-in-artificial-intelligence/

  [3] https://www.exin.com/en/certifications/bc-foundation-certificate-in-artificial-intelligence/

  [4] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3859976/