4. Design Methodologies for AI Applications

Try these first

1. 1.

   What is supervised learning?

   Hint

   Think of how we learn from guides.

2. 2.

   Define overfitting.

   Hint

   Consider if you memorize answers for a test but don’t understand the concepts.

3. 3.

   What is the main purpose of model evaluation?

   • To improve data quality
   • To understand how well the model performs with new data
   • To increase training duration

   Hint

   Think about why we check our answers after a test.

4. 4.

   True or False: Unsupervised learning requires labeled data.

   • True
   • False

   Hint

   Consider the nature of the data used.

5. 5.

   Propose a full design methodology for an AI application aimed to improve customer service in a retail setting. Outline the problem definition, algorithm selection, data sourcing, and evaluation metrics.

   Hint

   Think critically about each step involved and how they connect.

6. 6.

   Analyze how improper data preprocessing might lead to model failure or inaccuracies. Provide a scenario where this could happen.

   Hint

   Use examples from previous discussions to illustrate your points.