Question 13 1 Point Principal component analysis is a dimension reducing technique that produces new components that are A linear combinations of the original variables and maximise the variance of the independent variables. B linear combinations of the original orthogonalized variables and maximise the variance explained are linear combinations that are orthogonal in space and maximise the variance explained in the response D none of the aboveQuestion 14 Distance Matrix A B C D E F 0 16 47 72 16 0 37 57 47 37 0 40 30 72 57 40 0 31 77 65 30 31 0 79 66 35 23 10 Look at the attached distance matrix. If this distance matrix was used in a hierarchical cluster analysis which cases would merge at the next iteration? !:! none ofthe above Question 15 1 Point . .. Scree Plot Eigenvalue 1 5 6 7 8 10 11 12 Component Number Look at the scree plot. Which of the following is true A This analysis has p=10 B The first two or three principal components would be a good representation of the x variable space C 7 principal components is optimal D none of the aboveQuestion 16 If an independent variable is categorical ie coded with 3 levels [A. B and C]. how is this handled in a multiple linear regression model? Explain in detail. [Hint: 3-4 sentences should be enough] Use the editor to format your answer Question 17 In the context of regression analysis. explain what the term 'heteroscedasticicy' means and how this would be checked. (Hint: 2- 3 sentences should be enough] Use the editor to format your answer Question 18 The lasso and ridge methods are two procedures used in a regression analysis. Describemways how these techniques differ. [Hint: 2-3 sentences should be enough] Use the editor to format your