4.2.2. Algorithm Selection and Model Design

Try these first

1. 1.

   What is supervised learning?

   Hint

   Think about how it learns from examples.

2. 2.

   Give an example of unsupervised learning.

   Hint

   What method finds patterns in unlabelled data?

3. 3.

   What defines supervised learning?

   • Uses labeled data
   • Uses unlabeled data
   • Does not learn patterns

   Hint

   Consider what type of data the model is learning from.

4. 4.

   True or False: Transfer learning can save time when training models.

   • True
   • False

   Hint

   Think about how reusing models can speed things up.

5. 5.

   Design a comprehensive strategy using ensemble methods for a classification problem in predicting customer churn.

   Hint

   Think about mixing the strengths of different model types.

6. 6.

   Create a scenario where transfer learning could significantly improve a model's performance, detailing your reasoning.

   Hint

   Consider reusing already learned features to save time and effort.