**Unformatted text preview: **AVB = 0 or BVA=0 given that the r th cumulant of x 'Ax is 2 r-1 (r-1)![tr(AV) r +r 'A(VA) r-1 ] 5. For x n 1 ~N( ,V), and k<n, show that the marginal density function of y 1 where y' 1 =(x 1 ,x 2 ,...,x k ) is multivariate normal. 6. State and prove loynes lemma. 7. For the full rank model y =X + , show that the total sum of squares can be expressed as a sum of quadratic forms and give the distribution of the ratio of each to 2 . 8. For the model of full rank, y =X + for which all assumptions are met, derive an expression for the vector of estimators, H ˆ under the general linear hypothesis H: K =m. 9. What is the difference between outliers and influential data points? 10. What are the indicators of multicollinearity? Page 1 of 1...

View
Full Document

- Fall '19
- student