1. Numerical Analysis 1 EE, NCKU Tien-Hao Chang (Darby Chang)

2. In In the previous slide Special matrices –Strictly diagonally dominant matrix –Symmetric positive definite matrix Cholesky decomposition –Tridiagonal matrix Iterative techniques –Jacobi, Gauss-Seidel and SOR methods –conjugate gradient method Nonlinear systems of equations Exercise 3 2

3. In this slide Eigenvalues and eigenvectors The power method –locate the dominant eigenvalue Inverse power method Deflation 3

4. Chapter 4 4 Eigenvalues and Eigenvectors

5. Eigenvalues and eigenvectors Eigenvalue – λ– λ – Av=λv  (A-λI)v=0– Av=λv  (A-λI)v=0 – det(A-λI)=0 characteristic polynomial Eigenvector –the nonzero vector v for which Av=λv associated with the eigenvalue λ 5

6. In Chapter 4 Determine the dominant eigenvalue Determine a specific eigenvalue Remove a eigenvalue 6

7. 4.1 7 The Power Method

8. The power method Different problems have different requirements –a single, several or all of the eigenvalues –the corresponding eigenvectors may or may not also be required To handle each of these situations efficiently, different strategies are required The power method –an iterative technique –locate the dominant eigenvalue –also computes an associated eigenvector –can be extended to compute eigenvalues 8

9. The power method Basics 9

10. 10

11. 11

12. The power method Approximated eigenvalue 12

13. Any Questions? 13

14. The power method A common practice Make the vector x(m) have a unit length Why we need this step? 14 question

15. Make the vector x(m) have a unit length –to avoid overflow and underflow 15

16. The power method Complete procedure 16

17. Any Questions? 17

18. 18 In action http://thomashawk.com/hello/209/1017/1024/Jackson%20Running.jpg

19. 19 what is the first estimate?

20. 20

21. 21

22. Any Questions? 22 The power method for generic matrices

23. 23 The power method for symmetric matrices

24. When A is symmetric –more rapid convergence still linear, but smaller asymptotic error –different scaling scheme (norm) –based on the theorem 24

25. The power method variation 25

26. 26

27. Any Questions? 27 4.1 The Power Method

28. 28 An application of eigenvalue

29. 29

30. 30

31. Undirected graph Relation to eigenvalue Proper coloring –how to color the geographic regions on a map regions that share a common border receive different colors Chromatic number –the minimum number of colors that can be used in a proper coloring of a graph 31

32. Undirected graph The dominant eigenvalue 32

33. Undirected graph The corresponding eigenvector 33

34. Any Questions? 34

35. 4.2 35 The Inverse Power Method

36. The inverse power method To find an eigenvalue other than the dominant one To derive the inverse power method, we will need –the relationship between the eigenvalues of a matrix A to a class of matrices constructed from A With that, we can –transform an eigenvalue of A the dominant eigenvalue of B – B=(A-qI) -1 36 later

37. B is a polynomial of A 37

38. 38

39. 39

40. The inverse power method To find an eigenvalue other than the dominant one To derive the inverse power method, we will need –the relationship between the eigenvalues of a matrix A to a class of matrices constructed from A With that, we can –transform an eigenvalue of A the dominant eigenvalue of B – B=(A-qI) -1 40

41. 41

42. Any Questions? 42

43. How to 43 Find the eigenvalue smallest in magnitude

44. Any Questions? 44 4.2 The Inverse Power Method

45. 4.3 45 Deflation

46. So far, we can approximate –the dominant eigenvalue of a matrix –the one smallest in magnitude –the one closest to a specific value What if we need several of the largest/smallest eigenvalues? Deflation –to remove an already determined solution, while leaving the remainder solutions unchanged 46

47. Within the context of polynomial rootfinding –remove each root by dividing out the monomial – x 3 -6x 2 +11x-6 = (x-1)(x 2 -5x+6) = (x-1)(x-2)(x-3) – x 2 -5x+6 = (x-2)(x-3) is a deflation of x 3 -6x 2 +11x-6 For the matrix eigenvalue problem –shift the previously determined eigenvalue to zero (while leaving the remainder eigenvalues unchanged) –to do this, we need the relationship among the eigenvalues of a matrix A and A T 47

48. 48

49. 49 http://www.dianadepasquale.com/ThinkingMonkey.jpg Recall that

50. Deflation Shift an eigenvalue to zero 50

51. 51 While leaving the remainding eigenvalues unchanged

52. 52

53. 53

54. Deflation Summary 54

55. Any Questions? 55

56. Do we 56 Miss something?

57. 57 http://www.dianadepasquale.com/ThinkingMonkey.jpg Recall that

58. 58 ?How to choose x for the formula B=A-λ 1 v 1 x T ?

59. Wielandt Deflation 59

60. 60

61. Wielandt deflation Bonus 61

62. 62 In action http://thomashawk.com/hello/209/1017/1024/Jackson%20Running.jpg

63. 63

64. 64

65. Hotelling deflation Recall that we choose v 1,k based on infinity norm Like the power method, there is another deflation variation for symmetric matrices 65

66. 66

67. Any Questions? 67 4.3 Deflation