MATH 685 CSI 700 OR 682 Lecture Notes

1
MATH 685/ CSI 700/ OR 682 Lecture Notes

• Lecture 4.
• Least squares

2
Method of least squares

• Measurement errors are inevitable in
  observational and experimental sciences
• Errors can be smoothed out by averaging over
  many
• cases, i.e., taking more measurements than are
  strictly
• necessary to determine parameters of system
• Resulting system is overdetermined, so usually
  there is no exact solution
• In effect, higher dimensional data are projected
  into lower
• dimensional space to suppress irrelevant detail
• Such projection is most conveniently
  accomplished by
• method of least squares

3
Linear least squares
4
Data fitting
5
Data fitting
6
Example
7
Example
8
Example
9
Existence/Uniqueness
10
Normal Equations
11
Orthogonality
12
Orthogonality
13
Orthogonal Projector
14
Pseudoinverse
15
Sensitivity and Conditioning
16
Sensitivity and Conditioning
17
Solving normal equations
18
Example
19
Example
20
Shortcomings
21
Augmented system method
22
Augmented system method
23
Orthogonal Transformations
24
Triangular Least Squares
25
Triangular Least Squares
26
QR Factorization
27
Orthogonal Bases
28
Computing QR factorization

• To compute QR factorization of m n matrix A,
  with m gt n, we
• annihilate subdiagonal entries of successive
  columns of A,
• eventually reaching upper triangular form
• Similar to LU factorization by Gaussian
  elimination, but use
• orthogonal transformations instead of
  elementary elimination matrices
• Possible methods include
• Householder transformations
• Givens rotations
• Gram-Schmidt orthogonalization

29
Householder Transformation
30
Example
31
Householder QR factorization
32
Householder QR factorization
33
Householder QR factorization

• For solving linear least squares problem, product
  Q of
• Householder transformations need not be formed
  explicitly
• R can be stored in upper triangle of array
  initially
• containing A
• Householder vectors v can be stored in (now zero)
  lower
• triangular portion of A (almost)
• Householder transformations most easily applied
  in this
• form anyway

34
Example
35
Example
36
Example
37
Example
38
Givens Rotations
39
Givens Rotations
40
Example
41
Givens QR factorization
42
Givens QR factorization

• Straightforward implementation of Givens method
  requires
• about 50 more work than Householder method, and
  also
• requires more storage, since each rotation
  requires two
• numbers, c and s, to define it
• These disadvantages can be overcome, but requires
  more
• complicated implementation
• Givens can be advantageous for computing QR
• factorization when many entries of matrix are
  already zero,
• since those annihilations can then be skipped

43
Gram-Schmidt orthogonalization
44
Gram-Schmidt algorithm
45
Modified Gram-Schmidt
46
Modified Gram-Schmidt QR factorization
47
Rank Deficiency

• If rank(A) lt n, then QR factorization still
  exists, but yields
• singular upper triangular factor R, and multiple
  vectors x
• give minimum residual norm
• Common practice selects minimum residual solution
  x
• having smallest norm
• Can be computed by QR factorization with column
  pivoting
• or by singular value decomposition (SVD)
• Rank of matrix is often not clear cut in
  practice, so relative
• tolerance is used to determine rank

48
Near Rank Deficiency
49
QR with Column Pivoting
50
QR with Column Pivoting
51
Singular Value Decomposition
52
Example SVD
53
Applications of SVD
54
Pseudoinverse
55
Orthogonal Bases
56
Lower-rank Matrix Approximation
57
Total Least Squares

• Ordinary least squares is applicable when
  right-hand side b is subject to random error but
  matrix A is known accurately
• When all data, including A, are subject to error,
  then total least squares is more appropriate
• Total least squares minimizes orthogonal
  distances, rather than vertical distances,
  between model and data
• Total least squares solution can be computed from
  SVD of A, b

58
Comparison of Methods
59
Comparison of Methods
60
Comparison of Methods