Linear Regression and Model Evaluation Questions

What is a residual in linear regression?

• The slope of the regression line

• The difference between the actual and predicted values

• The predicted value

• The standard deviation of the error

Which of the following is a sign of multicollinearity?

• Residuals increase with predicted values

• R² is very low

• Variance Inflation Factor (VIF) is high for some features

• Mean squared error is negative

Which encoding method is generally preferred for categorical variables in linear regression?

• Label encoding

• One-hot encoding

• Binary encoding

• Hash encoding

Which of these is NOT an assumption of linear regression?

• A) Linear relationship between features and target

• B) Errors are normally distributed

• C) Features must be on the same scale

• D) Homoscedasticity of residuals

What is homoscedasticity?

• When errors are correlated

• When residuals have constant variance

• When features have high collinearity

• When features are normally distributed

What does a high R² score generally indicate?

• The model has underfit

• The model fits the data well

• Multicollinearity is present

• The residuals are large

What problem can arise if categorical variables are label encoded for linear regression?

• It increases accuracy

• The model interprets the values as ordinal

• It improves interpretability

• No issues will occur

Which metric is not appropriate for evaluating a linear regression model?

• Mean Squared Error (MSE)

• R² Score

• Accuracy

• Mean Absolute Error (MAE)