Module 4: Linear Classification

2. Graded Quiz: Linear Classification

1. Which option lists the steps of training a logistic regression model in the correct order?

Steps:

1. Use the cost function on the training set.

2. Update weights with new parameter values.

3. Calculate cost function gradient.

4. Initialize the parameters.

5. Repeat until specified cost or iterations reached.

Options: A. 1, 4, 3, 2, 5

B. 3, 2, 5, 4, 1

C. 4, 3, 2, 5, 1

D. 4, 1, 3, 2, 5

Answer: D. 4, 1, 3, 2, 5

2. What is the objective of SVM in terms of hyperplanes?

A. Minimize the distance between hyperplane and the support vectors.

B. Choose the hyperplane that represents the largest margin between the two classes.

C. Find the hyperplane of the lowest dimension.

D. Choose the hyperplane that’s closest to one of the two classes.

Answer: B. Choose the hyperplane that represents the largest margin between the two classes.

3. Logistic regression is used to predict the probability of a:

A. Numerical dependent variable

B. Categorical dependent variable

C. Numerical independent variable

D. Categorical independent variable

Answer: B. Categorical dependent variable

4. In which cases would we want to consider using SVM?

A. When we desire efficiency when using large datasets.

B. When mapping the data to a higher dimensional feature space can better separate classes.

C. When we want multiple decision boundaries with varying weights.

D. When we desire probability estimates for each class.

Answer: B. When mapping the data to a higher dimensional feature space can better separate classes.

5. What is a disadvantage of one-vs-all classification?

A. It requires more models to be created compared to one-vs-one.

B. There’s an ambiguous region where multiple classes are valid outputs.

C. It cannot output probability estimates of classes.

D. It does not handle two-class classification well.

Answer: B. There’s an ambiguous region where multiple classes are valid outputs.