CAIC 7

Question 1

What should you always keep in mind when using ChatGPT?

Answer 1

The limitations mentioned in previous chapters, as ChatGPT may provide partial or incorrect information.

It is a good practice to double-check the information provided.

Question 2

What type of questions should be avoided when using ChatGPT?

Answer 2

Vague, open-ended questions.

Examples include ‘What can you tell me about the world?’ or ‘Can you help me with my exam?’

Question 3

What is a best practice when expecting a specific output structure from ChatGPT?

Answer 3

Specify that structure in your prompt.

Question 4

What is the knowledge base limit of ChatGPT?

Answer 4

Limited to 2021.

Question 5

What is the purpose of the Moderator API in ChatGPT?

Answer 5

To prevent engagement in unsafe conversations.

Question 6

What are the classes used by the Moderator API to classify content?

Answer 6

• Violence
• Self-harm
• Hate
• Harassment
• Sex

Question 7

What is the hidden bias in GPT-3’s training data attributed to?

Answer 7

Mainly written by white males from Western countries.

Question 8

What did the study by OpenAI researchers reveal about racial bias in GPT-3?

Answer 8

Sentiment associated with racial categories varied across different models.

Question 9

What does the concept of responsible AI encompass?

Answer 9

Bias and ethics within AI models.

Question 10

What is the historical evolution of machine learning (ML)?

Answer 10

From checker game-playing programs in the 1950s to advanced AI like ChatGPT.

Question 11

What significant change has occurred in the technology infrastructure for ML?

Answer 11

Evolved from single machine/server to complex end-to-end ML platforms.

Question 12

What new professional roles have emerged due to hyper-growth in AI/ML?

Answer 12

• ML Engineers
• Data Scientists
• AI Ethics Researchers
• Data Analysts
• AI Product Managers

Question 13

What is the role of an ML solutions architect?

Answer 13

To support end-to-end ML initiatives.

Question 14

What is the first stage in the ML lifecycle?

Answer 14

Business understanding.

Question 15

What must be defined to measure the success of an ML project?

Answer 15

Business goals and business metrics.

Question 16

What is a common business goal for ML projects?

Answer 16

Cost reduction for operational processes.

Question 17

What does the saying ‘data is the new oil’ imply in the context of ML?

Answer 17

The necessity of having the required data to move forward with an ML project.

Question 18

What is involved in the data acquisition and understanding stage of the ML lifecycle?

Answer 18

Gathering and comprehending available data.

Question 19

What is feature engineering?

Answer 19

The process of using domain knowledge to extract useful features from raw data.

Question 20

What must be validated before deploying a model into production?

Answer 20

Model quality using relevant technical metrics.

Question 21

What is a validation dataset also known as?

Answer 21

Test dataset.

Question 22

Why is model accuracy not always a suitable validation metric?

Answer 22

It may not reflect performance well in cases like fraud detection where the number of frauds is small.

Question 23

What type of project structure is typical in an ML project?

Answer 23

• Business understanding
• Data acquisition and understanding
• Data preparation
• Model building
• Model evaluation
• Model deployment

Question 24

What was the author’s previous experience before working in AI/ML?

Answer 24

Building computer software platforms for large financial services institutions.

Question 25

What does the iterative process in ML involve?

Answer 25

Numerous runs of data processing and model development to find optimal performance.

Question 26

What challenges did the author face in deploying the model?

Answer 26

Integrating it into the existing business workflow and system architecture.

Question 27

What is essential to ensure before proceeding with an ML project?

Answer 27

Sufficient justification and measurable outcomes.

Question 28

What is the purpose of model validation in machine learning?

Answer 28

To gauge how the model performs on unseen data ## Footnote Model validation ensures that the model generalizes well beyond the training dataset.

Question 29

What factors determine the appropriate metrics for model validation?

Answer 29

ML problems and the dataset used ## Footnote Different problems and datasets require different evaluation metrics.

Question 30

Why would model accuracy not be a good metric for evaluating fraud detection models?

Answer 30

The number of frauds is small, resulting in potentially high accuracy despite poor performance ## Footnote A model predicting not-fraud all the time could still achieve high accuracy.

Question 31

What are the two main deployment concepts in machine learning?

Answer 31

Deployment of the model for client applications and integration into business workflow applications ## Footnote These concepts ensure that the model's predictions are utilized effectively.

Question 32

How can a credit fraud model be deployed?

Answer 32

Hosted behind an API for real-time prediction or as a package for batch predictions ## Footnote This allows flexibility in how predictions are generated and used.

Question 33

What is a key post-deployment step in the ML lifecycle?

Answer 33

Model monitoring ## Footnote Monitoring is crucial for detecting performance degradation and changes in data distribution.

Question 34

What is model drift?

Answer 34

Model performance degradation due to changes in production data characteristics ## Footnote This phenomenon can significantly impact the effectiveness of deployed models.

Question 35

What should be tracked to measure the actual business impact of a deployed model?

Answer 35

Business metrics before and after model deployment ## Footnote This helps in assessing the model's effectiveness and overall impact.

Question 36

What is A/B testing in the context of model evaluation?

Answer 36

Comparing business metrics between workflows with and without the ML model ## Footnote A/B testing helps determine the model's contribution to business outcomes.

Question 37

What should be done if a deployed model does not deliver expected benefits?

Answer 37

Re-evaluate the model for improvement opportunities or consider framing the problem differently ## Footnote This may involve exploring alternative ML approaches to address the business problem.

Question 38

Fill in the blank: The ML lifecycle does not end with _______.

Answer 38

[model deployment]

Question 39

True or False: Software behavior is highly deterministic while ML models can behave differently in production.

Answer 39

True ## Footnote This difference arises because ML models learn from data rather than being explicitly coded.