Logistic Regression: Test Your Knowledge

Let's do a quick test! You must answer at least 4 questions correctly to pass this quiz.

1. Why is a linear regression model's output a poor predictor of probability?

   It only has one weight per feature.

   It only has one output value.

   The bias parameter skews the output value.

   Its predictions are not restricted to values between 0 and 1.

2. True or false: A sigmoid function never outputs the value 0 or the value 1.

   True

   False

3. True or false: Applying regularization is less important when training logistic regression models than it is for training linear regression models.

   True

   False

4. Which of the following options matches both Linear Regression and Logistic Regression with appropriate loss functions for calculating loss?

   Linear Regression: Mean squared error
   Logistic Regression: Mean squared error

   Linear Regression: Mean squared error
   Logistic Regression: Mean absolute error

   Linear Regression: Mean squared error
   Logistic Regression: Log Loss

   Linear Regression: Log Loss
   Logistic Regression: Mean squared error

5. Which of the following is an effective regularization technique for logistic regression models?

   Dropout regularization

   Late stopping

   Early stopping

   Gradient descent

Submit answers

Was this helpful?