What is the difference between homoscedasticity and heteroscedasticity in regression, and why does it matter?

Machine Learning Hard

Machine Learning — Hard

What is the difference between homoscedasticity and heteroscedasticity in regression, and why does it matter?

Key points

  • Homoscedasticity: constant residual variance
  • Heteroscedasticity: varying residual variance
  • OLS assumption: homoscedasticity for accurate estimates
  • Inefficient estimates with heteroscedasticity
  • Incorrect standard errors with heteroscedasticity

Ready to go further?

Start full exam Practice more questions

Related questions