Discussion of Lectures 5

Xijia Liu

Department of Statistics, Umeå University

Lecture 5: Regression Models

  • Do you remember the names of prediction and prediction errors in the regression analysis?
  • MSE V.S. Likelihood function V.S. Loss function
  • Can you describe the basic idea of nonlinear extension in simple words?
  • What are the main challenges of this idea?
  • Can you explain what is overfitting problem?
  • What are the main reasons?
  • How do you interpret overfitting problem from variance-bias trade off point of view?