Practice Overfitting And Underfitting (12.6) - Evaluation Methodologies of AI Models

Practice: Overfitting and Underfitting

Lesson 12.6Evaluation Methodologies of AI Models

Practise this section

9 questions

Flashcard drill
4 cards from this section. Quick, good the night before a test.
Exercise
6 questions · about 4 min · pass mark 70%. Hints available.
Sign up to start the exercise
Quiz
3 questions · about 2 min · pass mark 70%. Hints available.
Sign up to start the quiz

6 sample questions

Question 1.
What happens when a model overfits?
Show hintHide hint
Think about how well a student does on practice tests versus the actual exam.
Question 2.
What is the main issue with underfitting?
Show hintHide hint
Consider how a student might miss key concepts in their studies.
3Question 3.
What is overfitting?
- A model that performs well on seen data only
- A model that performs equally well on new data
- A model that is too simple
Show hintHide hint
Consider how well the model can predict on new data versus training data.
4Question 4.
True or False: Underfitting refers to a model that has high variance.
- True
- False
Show hintHide hint
Remember the definitions of variance and bias.
Question 5.
Consider a dataset where the relationship between variables is quadratic, but you use a linear regression model. What issues arise, and how can this be resolved?
Show hintHide hint
What additional terms could capture the relationship between your variables?
Question 6.
You have data that is very noisy, and a model is trained to fit every point exactly. Explain the likely outcomes and the adjustments you could make to improve generalization.
Show hintHide hint
What techniques can you think of to penalize complexity?

5 more questions in this section. Enroll free to practise them all.