Transparent AI Exercises (Part 1)

Quiz Summary

0 of 10 Questions completed

Questions:

You have already completed the quiz before. Hence you can not start it again.

Quiz is loading…

You must sign in or sign up to start the quiz.

You must first complete the following:

Results

Quiz complete. Results are being recorded.

Results

0 of 10 Questions answered correctly

Your time:

You have reached 0 of 0 point(s), (0)

Current
Review
Answered
Correct
Incorrect

Question 1 of 10

1. Question
Define AI System Transparency: (Select the best answer)
- Providing algorithmic explanations and interpretations on how AI systems arrived at specific decisions
- Visibility into all the aspects of what went into building an AI system so users can understand the full context of how an AI system is built and used
- Visibility into the decision-making on implementation of AI systems
- Visibility into organizational change management considerations for AI and data system implementation
Correct

Incorrect
Question 2 of 10

2. Question
AI systems should provide visibility into: (select all that apply)
- Training and inference data used
- The components of the system with their configuration
- The human operator
- AI decision log files
- Setting and hyperparameter decisions for AI systems
- Procurement decisions for vendor tooling
Correct

Incorrect
Question 3 of 10

3. Question
What aspects of AI system Transparency are important for trust? (select all that apply)
- Understanding how the AI models are being built
- Understanding what data has been used to create the models
- Understanding what data was being included / excluded and reasons for their inclusion or exclusion
- Understanding how algorithms function
- Understanding how data science teams work
- Understanding how bias is measured and current bias measures
Correct

Incorrect
Question 4 of 10

4. Question
When creating your Trustworthy AI framework, you need to consider transparency around the following: (select all that apply)
- Training Data selection
- Configuration and parameter settings
- Algorithm selection
- Automation decisions
- Vendor selection
- Human decisions relevant to the AI system
Correct

Incorrect
Question 5 of 10

5. Question
When creating your Trustworthy AI framework, you need to address the following specific transparency requirements: (select all that apply)
- How will you provide visibility into training data used?
- How will you provide visibility into potential use cases?
- How will you document and share methods of data selection?
- What approaches will you use to provide greater visibility into the various decisions and design considerations made for your AI systems?
- What approaches will you use to provide greater visibility into root cause analysis?
- How will you provide visibility into key performance indicators and ROI metrics?
Correct

Incorrect
Question 6 of 10

6. Question
Human decisions should be open and transparent around the following: (select all that apply)
- Operation of AI systems
- Versioning of ML models and AI systems
- The seven patterns of AI
- Organizational team alignment
- AI development
- Applications and use of the AI system
Correct

Incorrect
Question 7 of 10

7. Question
Bias Measurement & Mitigation should be addressed in the following ways for your Trustworthy AI framework: (select all that apply)
- Addressing informational bias
- Addressing selection bias
- Addressing development bias
- AI systems should provide a means to constantly measure bias of various sources
- How to provide means to mitigate any bias detected
- Limiting access to data
Correct

Incorrect
Question 8 of 10

8. Question
When addressing AI System Disclosure in your Trustworthy AI framework you need to make sure to address the following: (select all that apply)
- How (if at all) you disclose to your users that an AI system is being used
- When you will disclose to your users that an AI system is being used
- Ways in which to use “Pseudo-AI”
- Methods to disguise use of AI in applications
- Emerging laws and regulations on AI disclosure
- How and when you disclose that a human might be in the loop
Correct

Incorrect
Question 9 of 10

9. Question
What approaches can you use to mitigate bias in the Trustworthy AI framework: (select all that apply)
- Provide Approaches such as failovers, fallbacks, non-AI approaches when systems give low levels of confidence in data sets with high bias
- Provide ways to automate decision making when bias is identified
- Provide alternatives for high-risk decision making when bias is identified
- Address how you will communicate bias issues to internal and external parties
- Address how you will identify and address variance in datasets
- Reduce access to data sets
Correct

Incorrect
Question 10 of 10

10. Question
AI systems should use open source technology when possible for the following reasons: (select all that apply)
- There is a very large, growing pool of open training data sets that others have contributed to
- Internal models and data are often better than open data sets
- Open source data sets and technology are usually implemented in the Cloud
- Open AI data sources provide visibility into the training data and design choices
- Open Source Communities are constantly improving and inspecting, fixing the data
- Vendor-provided, Third party data or models often provide limited transparency
Correct

Incorrect

Transparent AI Exercises (Part 1)

Transparent AI Exercises (Part 1)

Quiz Summary

Information

Results

Results

Categories

1. Question

2. Question

3. Question

4. Question

5. Question

6. Question

7. Question

8. Question

9. Question

10. Question