Module 3: Understanding the AI Development Lifecycle: Plan

Question 1

What are the phases of the AI Development Lifecycle?

Answer 1

• Planning
• Design
• Development
• Implementation

Question 2

What are the steps in the Planning phase?

Answer 2

• Determine the business objectives and requirements.
• Determine the scope of the project.
• Determine the governance structure and responsibilities.

Question 3

Name three main types of business problems.

Answer 3

1) Classification
2) Regression (predict the future based on past data)
3) Recommendation

Question 4

What are key questions to ask when identifying the business problem (business objectives and requirements)?

Answer 4

• What results do you expect to see?
• What processes are currently in use?
• What type of improvement are you expecting?
• What are the KPIs?
• What resources are available?

Question 5

What are some things that can be done to assist with identifying the business problem?

Answer 5

• Conduct user interviews
• Conduct market research on AI systems
• Identify AI use cases
• Identify the gaps
• Ensure you use the right data

Question 6

What are some steps to take when determining the scope of the project?

Answer 6

• Prioritize problems to solve

Question 7

What are the 3 main elements to look at when prioritizing which problems to solve?

Answer 7

1) Impact of use of AI system
2) Effort needed (timeframe & resources)
3) Fit of AI system

Question 8

What are the steps in determining the governance structure and responsibilities?

Answer 8

• Determine the governance structure (if any)
• Who is responsible for maintaining the AI Governance structure?
• Identify a champion for development/implementation

Question 9

What steps do you need to take when designing the AI development lifecycle?

Answer 9

• Implement a data strategy that includes: data gathering, wrangling, cleansing, labeling; applying PETs like anonymization, minimization, differential privacy, federated learning.
• Determine AI system architecture and model selection (choose the algorithm according to the desired level of accuracy and interpretability).

Question 10

What questions should be asked during data gathering?

Answer 10

• What data is required?
• How much data is needed?
• How is data collected and stored?
• Do you want to use pre-trained data?
• Do you want to include external data?
• Consider the quality of the data
• Consider the format of the data

Question 11

What are some different formats of data?

Answer 11

• Structured data (labeled)
• Unstructured data (e.g. images)
• Static data (e.g. past sales)
• Streaming data (data that will change - e.g. customers visiting a website)