Logistic Regression

Why do we use logistic regression instead of linear regression?

• Logistic regression fits a linear equation to our data!

• It is more logical

• Logistic regression classifies our inputs!

• Logistic regression is supervised learning, and linear regression is not.

What is the loss for multiclass classification?

• Squared error

• 
• Absolute error

• 

What is the loss for binary classification?

• Absolute error

• 
• Squared error

• 

What does the sigmoid / softmax function do?

• Makes sure the output is continuous, from -∞ to ∞

• Alters the output to be in the range [0, 1]

• Maps predictions to probabilities.

• Sigmoid is linear

What do we call the line that separates classes?

• Linear regression

• Decision boundary

• Loss function