Student: Kyle Hoang Date: 04/15/22 Instructor: Larissa Williamson Course: MAS3114-Assignment: M30 (HW) Approximation &

4/15/22, 1:20 PM M30(HW) Approximation & Gram-Schmidt Process (6.
3,-Kyle Hoang
Instructor: Larissa Williamson

Student: Kyle Hoang Course: MAS3114- Assignment: M30(HW) Approximation &
Date: 04/15/22 1888,0329(15272,28798) - Computatnl Gram-Schmidt Process (6.3,
Linear Alg
Find the closest point to y in the subspace W spanned by v1 and v2 .
− 13 1 6
−1 0 1
y= , v1 = , v2 =
1 −1 0
11 −2 3
Recall the Best Approximation Theorem. Let W be a subspace of ℝn , let y be any vector in ℝn , and let y be the orthogonal
projection of y onto W. Then y is the closest point in W to y, in the sense that y − y < y − v for all v in W distinct from y .
Since W = Span v1 ,v2 is a subspace of ℝn , the Best Approximation Theorem can be used to find the closest point to y in the
subspace W spanned by v1 and v2 for the given vectors.
Let W be a subspace of ℝn . Then each y in ℝn can be written uniquely in the form y = y + z where y is in W and z is in W . If
u1 , . . . , up is any orthogonal basis ofW, then
y • u1 y • up
y= u1 + • • • + u
u1 • u1 up • up p
and z = y − y . The vector y is called the orthogonal projection of y onto W.
Remember, for thisproblem, W = Span v1 ,v2 . To use the above definition of y , verify that v1 ,v2 is an orthogonal set. The
set v1 ,v2 is orthogonal if and only if v1 • v2 = 0.
Compute v1 • v2 .
u1 v1
u2 v2
T
If u and v are vectors in , with u = • and v = • , then u • v = u v,
• •
• •
un vn
or u • v = u1 v1 + u2 v2 + • • • + un vn .
Substitute the components of v1 and v2 into the expression for v1 • v2 .
v1 • v2 = (1)(6) + (0)(1) + ( − 1)(0) + ( − 2)(3)
Simplify v1 • v2 .
v1 • v2 = (1)(6) + (0)(1) + ( − 1)(0) + ( − 2)(3)

v1 • v2 = 0
Since v1 • v2 = 0, the set v1 ,v2 is orthogonal and the previously stated definition of y can be used.
https://xlitemprod.pearsoncmg.com/api/v1/print/highered 1/2
4/15/22, 1:20 PM M30(HW) Approximation & Gram-Schmidt Process (6.3,-Kyle Hoang
By thatdefinition, the formula that should be used to find y , the orthogonal projection of y onto Span v1 ,v2 , is
y • v1 y • v2
y= v1 + v .
v1 • v1 v2 • v2 2
Use this formula to find y . Begin by finding the four inner products.
y • v1 y • v2
y= v1 + v
v1 • v1 v2 • v2 2
− 36 − 46
y= u1 + u2
6 46
Substitute for v1 and v2 and simplify the resulting expression.
1 6
− 36 0 − 46 1
y= +
6 −1 46 0
−2 3
− 12
−1
y=
6
9
− 12
−1
Thus, the closest point to y in the subspace W spanned by v1 and v2 is the vector .
6
9

Linear Alg
Find the best approximation to z by vectors of the form c1 v1 + c2 v2 .
8 9 1
−5 6 −1
z= , v1 = , v2 =
3 0 12
8 −3 1
Let W be a subspace of ℝn , let y be any vector in ℝn , and let y be the orthogonal projection of y onto W.Then, y is the closest
point in W to y in the sense that y − y < y − v for all v distinct from y .
Let y exist in ℝn and let W be a subspace of ℝn . Then, if u1 ,...,up is any orthogonal basis ofW,
y • u1 y • up
y= u1 + ⋯ + u is the orthogonal projection of y onto W.
First, notice that z, v1 , and v2 exist in ℝ4 . Now, let W = Span{v1 , v2 }. Since v1 and v2 exist in ℝ4 , W is a subspace of ℝ4 . To
find the orthogonal projection of y ontoW, it is necessary to find an orthogonal basis for W.
Consider {v1 , v2 }. Since Span{v1 , v2 } coincides with W and c1 v1 + c2 v2 = 0 has one unique nontrivial solution of c1 = c2 = 0, it
is clear that {v1 , v2 } is a basis of W. To show that {v1 , v2 } is an orthogonalbasis, it is necessary to show that {v1 , v2 } is an
orthogonal set.
For {v1 , v2 } to be an orthogonalset, the inner products of each pair of elements must equal 0.
Compute the appropriate inner products. Find v1 • v2 .
v1 • v2 = 0
Since v1 • v2 = 0, the inner product is 0 and the set {v1 , v2 } is orthogonal.Thus, {v1 , v2 } is an orthogonal basis of W.
z • v1 z • v2 z • v1
So, the orthogonal projection of z onto W is z such that z = v1 + v2 . To find z , start by computing .
v1 • v1 v2 • v2 v1 • v1
Find z • v1 .
z • v1 = 18
Now, find v1 • v1 .
v1 • v1 = 126
z • v1 1
Thus, = .
v1 • v1 7
z • v2
Now, compute similarly.
v2 • v2
z • v2 19
=
v2 • v2 49
9 1
1 6 19 −1
So, z = + .
7 0 49 12
−3 1
Multiply and add to find z .
82
49
23
49
z=
228
49
2
−
49
Thus, by the Best ApproximationTheorem, the best approximation of z is as follows.
82
49
9 1 23
1 6 19 −1 49
z= + , or z =
7 0 49 12 228
−3 1 49
2
−
49

Linear Alg
2 −3 −1
3
Let y = − 4 , u1 = − 5 , u2 = 1 . Find the distance from y to the plane in ℝ spanned by u1 and u2 .
7 1 2
Note that the distance from a point y in ℝ3 to a subspace W is defined as the distance from y to the closest point in W.
Since the closest point in W to y is y = projWy, the desired distance is y − y . Calculate y with the following theorem.
Let W be a subspace of ℝn . Then each y in ℝn can be written uniquely in the form y = y + z where y is in W and z is in W . If
⊥
u1 , ..., up is any orthogonal basis ofW, then
y • u1 y • up
y= u1 + ⋯ + u
and z = y − y . The vector y is called the orthogonal projection of y onto W.
y • u1 y • u2
To find y , use the formula y = u1 + u .
u1 • u1 u2 • u2 2
Use this formula to find y . Begin by finding y • u1 .
y • u1 y • u2
y= u1 + u
u1 • u1 u2 • u2 2
−3 y • u2 −1
21
y= −5 + 1
u1 • u1 u2 • u2
1 2
Next, find u1 • u1 .
−3 y • u2 −1
21
y= −5 + 1
u1 • u1 u2 • u2
1 2
−3 y • u2 −1
21
y= −5 + 1
35 u2 • u2
1 2
Now, find y • u2 .
−3 −1
21 y • u2
y= −5 + 1
35 u2 • u2
1 2
−3 −1
21 8
y= −5 + 1
35 u2 • u2
1 2
Find u2 • u2 .
−3 −1
21 8
y= −5 + 1
35 u2 • u2
1 2
−3 −1
21 8
y= −5 + 1
35 6
1 2
Finally, simplify the expression.
47
−
15
−3 −1
21 8 5
y= −5 + 1 = −
35 6 3
1 2
49
15
47
−
15
5
Thus, y = − .
3
49
15
Calculate y − y .
47 77
−
15 15
2
5 7
y−y = −4 − − = −
3 3
7
49 56
15 15
77
15
7
Therefore, y − y = − .
3
56
15
To calculate y − y , recall the following theorem. The length(or norm) of v is the nonnegative scalar v defined as follows.
2 2 2
v = v•v = v1 + v2 + ⋯ + vn
Find y − y .
2 2 2
y−y = y−y 1 + y−y 2 + y−y 3
2 2 2
77 7 56
= + − +
15 3 15
14
= 7
15
14
Therefore, the distance from y to the plane in ℝ3 spanned by u1 and u2 is 7 .
15

Linear Alg
− 13 1 6
−1 0 1
Let y = , v1 = , and v2 = . Find the distance from y to the subspace W of ℝ4 spanned by v1 and v2 , given
1 −1 0
11 −2 3
− 12
−1
that the closest point to y in W is y = .
6
9
The distance from a point y in ℝn to a subspace W is defined as the distance from y to the nearest point in W.
Recall that the closest point to y in W is y . By the definition of the distance between two vectors in ℝn , the distance from y to y
is given by y − y .
Compute y − y .
− 13 − 12 −1
−1 −1 0
y−y= − =
1 6 −5
11 9 2
2 2 2 2
By the definition of the length of a vector in ℝn , y − y is given by the expression ( − 1) + 0 + ( − 5) + 2 .
Compute y − y .
2 2 2 2
y−y = ( − 1) + 0 + ( − 5) + 2
y−y = 30
Therefore, the distance from y to the subspace W of ℝ4 spanned by v1 and v2 , given that the closest point to y in W is
− 12
−1
y= , is 30 .
6
9

Linear Alg
4
19 17
Let y = , u1 = , and W = Span u1 . Complete parts(a) and(b) below.
8 1
−
17
a. Let U be the 2 × 1 matrix whose only column is u1 . Compute UT U and UUT .
T
Find the transpose ofU, U .
4 1
UT = −
17 17
Compute UT U.
T 4 1 17
U U= −
17 17 1
−
17
T
U U= 1
T
Compute UU .
4
17 4 1
UUT = −
1 17 17
−
17
16 4
−
17 17
UUT =
4 1
−
17 17
16 4
−
17 17
Therefore, UT U = I and UUT = .
4 1
−
17 17
T
b. Compute projW y and UU y.
n
Use the theorem which says that if u1 ,...,up is an orthonormal basis for a subspace W of ℝ , and if
T
U= u1 u2 • • • up , then projW y = UU y for all y in ℝn .
A set u1 ,...,up is an orthonormal set if it is an orthogonal set of unit vectors. If W is the subspace spanned by such aset, then
u1 ,...,up is an orthonormal basis for W.
For thisproblem, W = Span u1 , so W is a subspace of ℝn . Also, it was given in part(a) to let U be the 2 × 1 matrix whose only
column is u1 , so U = u1 .
Use the theorem which says that an m × n matrix U has orthonormal columns if and only if UT U = 1. From part(a), U = u1 ,
and it was found that UT U = 1. Thus, U has orthonormal columns. Since u1 is the only column ofU, it follows that the set u1
is an orthonormal basis for W = Span u1 .
T
With this result and the theorem stated at the beginning of part(b), it is now known that projW y = UU y. Recall from part(a)
16 4
−
17 17
that UUT = .
4 1
−
17 17
T
Compute projW y and UU y.
16 4
−
T 17 17 19
projW y = UU y=
4 1 8
−
17 17
T 16
projW y = UU y=
−4
T 16
Thus, projW y = UU y= .
−4

Linear Alg
−1 1 0
Let u1 = − 1 , u2 = − 2 , and u3 = 0 . Note that u1 and u2 are orthogonal but that u3 is not orthogonal to u1 or u2 . It
1 −1 1
can be shown that u3 is not in the subspace W spanned by u1 and u2 . Use this fact to construct a nonzero vector v in ℝ3 that
is orthogonal to u1 and u2 .
Let W be a subspace of ℝn . Then each y in ℝn can be written uniquely in the form y = y + z where y is in W and z is in W . In
⊥
fact, if u1 ,...,up is any orthogonal basis ofW, then y is given by the formula below and z = y − y .
y • u1 y • up
y= u1 + ⋯ + u
This exercise asks for a nonzero vector that is orthogonal to u1 and u2 . One such vector is the vector z in the theorem given
u3 • u1 u3 • u2
above. Thatis, one possible vector is v = u3 − u3 , where u3 = u + u .
u1 • u1 1 u2 • u2 2
To find u3 , first compute the inner products. Compute u3 • u1 .
u3 • u1 = (0)( − 1) + (0)( − 1) + (1)(1)

u3 • u1 = 1
Compute u1 • u1 .
u1 • u1 = ( − 1)( − 1) + ( − 1)( − 1) + (1)(1)

u1 • u1 = 3
Compute u3 • u2 .
u3 • u2 = (0)(1) + (0)( − 2) + (1)( − 1)

u3 • u2 = − 1
Compute u2 • u2 .
u2 • u2 = (1)(1) + ( − 2)( − 2) + ( − 1)( − 1)

u2 • u2 = 6
Substitute these values and the given vectors u1 and u2 into the formula for u3 .
u3 • u1 u3 • u2
u3 = u1 + u
u1 • u1 u2 • u2 2
−1 1
1 −1
u3 = −1 + −2
3 6
1 −1
Simplify this result.
1
−
−1 1 2
1 1
−1 − −2 = 0
3 6
1 −1 1
2
Substitute u3 and u3 into the formula for v.
v = u3 − u 3
1
−
0 2
v= 0 − 0
1 1
2
1 1
−
0 2 2
0 − 0 = 0
1 1 1
2 2
1
2
Therefore, a nonzero vector in ℝ3 that is orthogonal to u1 and u2 is v = 0 . Note that any multiple of v is also orthogonal to
1
2
u1 and u2 .

Linear Alg
7. Decide whether statements(a) through(e) are true or false. Justify each answer. Assume all vectors and subspaces are in
n
ℝ .
⊥
a. If z is orthogonal to u1 and u2 and if W = Span u1 ,u2 , then z must be in W . Choose the correct answer below.
A. The statement is true because W⊥ is the set of all vectors orthogonal to u and u , so by
1 2
⊥
definition, z is in W .
B. The statement is false because, since z is orthogonal to u1 and u2 , it exists in Span u1 ,u2 .
⊥
Since W = Span u1 ,u2 , z is in W and cannot be in W .
C. The statement is truebecause, since z is orthogonal to u1 and u2 , it is orthogonal to every

vector in Span u1 ,u2 , a set that spans W.
D. The statement is false because if z is orthogonal to u1 and u2 , it only follows that z orthogonal to
⊥
Span u1 and Span u2 . This is not enough information to conclude that z is in W .
b. For each y and each subspaceW, the vector y − projW y is orthogonal to W. Choose the correct answer below.
A. The statement is true because y can be written uniquely in the form y = projW y + z where
⊥
projW y is in W and z is in W and it follows that z = y − projW y.
B. The statement is false because y − projW y is in W and so cannot be orthogonal to W.
C. The statement is true because y and projW y are both orthogonal to W.Thus, a linear
combination of them must also be orthogonal to W.
D. The statement is false because y can be written uniquely in the form y = projW y + z where z is in
⊥
W and projW y is in W and it follows that z = y − projW y.
c. The orthogonal projection y of y onto a subspace W can sometimes depend on the orthogonal basis for W used to
compute y . Choose the correct answer below.
A. The statement is false because the uniqueness property of the orthogonal decomposition
y = y + z indicatesthat, no matter the basis used to findit, it will always be the same.
B. The statement is true because the orthogonal projection y of y onto a subspace W depends on
an orthonormal basis for W.
C. The statement is false because the orthogonal projection y of y onto a subspace W depends on
an orthonormal basis for W.
D. The statement is true because for each different orthogonalbasis, y is expressed as a different
linear combination of the vectors in that basis.
d. If y is in a subspaceW, then the orthogonal projection of y onto W is y itself. Choose the correct answer below.
A. The statement is false because if y is inW, then proj y is orthogonal to y, and is in W⊥ .
W
B. The statement is true because for an orthogonal basis ofW, B = u1 ,...,up , y and projW y can be
written as linear combinations of vectors in B with equal weights.
C. The statement is false because if y is inW, then projW y = 0, so this statement is false unless
y = 0.
D. The statement is true because if y is inW, then projW y = − y, which is in the same spanning set
as y.
T
e. If the columns of an n × p matrix U areorthonormal, then UU y is the orthogonal projection of y onto the column space of
U. Choose the correct answer below.
A. The statement is false because the columns of U are linearly independent and thus form an
T
orthogonal basis for Col U.So, projCol U y = U Uy for all y in ℝn .
B. The statement is true because the columns of U are linearly independent and thus form an
T
orthonormal basis for Col U.So, projCol U y = UU y for all y in ℝn .
C. The statement is false because the columns of U are linearly independent and thus form an
T
orthonormal basis for Col U.So, projCol U y = U Uy for all y in ℝn .
D. The statement is true because the columns of U are linearly independent and thus form an
T
orthogonal basis for Col U.So, projCol U y = UU y for all y in ℝn .

Linear Alg
8. All vectors and subspaces are in ℝn . Mark each statement True or False. Justify each answer. Complete parts(a) through
(e) below.
a. If W is a subspace of ℝn and if v is in both W and W , then v must be the zero vector.

⊥
True. If is inW, then Since the component of is equal to the component of must be A similar argument can be
formed for the W component of based on the orthogonal projection of onto the subspace . Thus, be .
y • u1 y • up
b. In the Orthogonal DecompositionTheorem, each term in y = u1 + ... + u is itself an orthogonal projection
of y onto a subspace of W.
True. Since is the projection of onto and the span of each is a one -dimensional subspace ofW, each resulting
projection must be onto the subspace spanned by
c. If y = z1 + z2 , where z1 is in a subspace W of ℝn and z2 is in W , then z1 must be the orthogonal projection of y onto W.

⊥
True. Since the orthogonal decomposition of into components that exist in W and is unique, must correspond to the
orthogonal projection of onto W.
d. The best approximation to y by elements of a subspace of W is given by the vector y − projW y.
False. The Best Approximation Theorem says that the best approximation to is
T
e. If an n × p matrix U has orthonormalcolumns, then UU x = x for all x in ℝn .
False. Let W be the subspace spanned by the columns of U. Then U. If pn, then W will be all of , so the statement is
true for all in . If pn, then W will not be all of , so the statement is not true for all in .

Linear Alg
The set is a basis for a subspace W. Use theGram-Schmidt process to produce an orthogonal basis for W. Assume the vectors
are in the order x1 , x2 .
6 −9
− 6 , 2
6 −7
The first step in theGram-Schmidt process is to define v1 = x1 .
6
v1 = −6
6
x2 • v1
The formula for v2 is x2 − v . Begin by finding the projection of x2 onto v1 .
v1 • v1 1
x2 • v1 ( − 9 • 6) + (2 • − 6) + ( − 7 • 6) − 108
= = = −1
v1 • v1 (6 • 6) + ( − 6 • − 6) + (6 • 6) 108
Complete the calculation of v2 .
−9 6 −9+6 −3
v2 = 2 − ( − 1) −6 = 2 + ( − 6) = −4
−7 6 −7+6 −1
6 −3
Therefore, the basis found usingGram-Schmidt is −6 , −4 .
6 −1
Verify this basis is orthogonal by showing that v1 • v2 = 0.
(6 • − 3) + ( − 6 • − 4) + (6 • − 1) = 0

Linear Alg
5 5 When given the orthogonalbasis, v1 , v2 , the orthonormal

basis is constructed by normalizing all the vectors in the
The vectors v1 = −6 and v2 = 3 form an orthogonal
orthogonal basis.
1 −7
1
basis for W. Find an orthonormal basis for W. The formula uk = vk shows how to normalize a vector vk
vk
of an orthogonal basis.
Begin by finding v1 to normalize v1 .
v1 = 62
Find u1 .
1
u1 = v1
v1
5
1
= −6
62
1
5
62
6
= −
62
1
62
Now find v2 to normalize v2 .
v2 = 83
Find u2 .
1
u2 = v2
v2
5
1
= 3
83
−7
5
83
3
=
83
7
−
83
The orthonormal basis of the subspace spanned by the

5 5
62 83
6 3
vectors is − , .
62 83
1 7
−
62 83

Linear Alg
Find an orthogonal basis for the column space of the matrix to the right. −1 5 7
3 −6 4
2 −1 7
1 −5 −4
Given a basis {x1 ,...,xp } for a nonzero subspace W of ℝn , define v1 ,...,vp as follows.
v1 = x1
x2 • v1
v2 = x2 − v
v1 • v1 1
x3 • v1 x3 • v2
v3 = x3 − v1 − v
v1 • v1 v2 • v2 2
⋯
xp • v1 xp • v2 xp • vp − 1
vp = xp − v − v −⋯− v
v1 • v1 1 v2 • v2 2 vp − 1 • vp − 1 p − 1
Then {v1 ,...,vp } is an orthogonal basis for W. This is called theGram-Schmidt process.
−1 5 7
3 −6 4
Let x1 = , x2 = , and x3 = . Notice that these vectors are linearly independent and hence form a basis for a
2 −1 7
1 −5 −4
n
nonzero subspace W of ℝ .
Since the vectors above form a basis for a nonzerosubspace, theGram-Schmidt process can be applied. Let v1 = x1 , so that
−1
3
v1 = . Use this and x2 to find v2 .
2
1
x2 • v1
Recall that v2 = x2 − v . Compute the inner products that appear in this formula. First compute x2 • v1 .
v1 • v1 1
x2 • v1 = (5)( − 1) + ( − 6)(3) + ( − 1)(2) + ( − 5)(1)

x2 • v1 = − 30
Next compute v1 • v1 .
v1 • v1 = ( − 1)( − 1) + (3)(3) + (2)(2) + (1)(1)

v1 • v1 = 15
Substitute − 30 for x2 • v1 , 15 for v1 • v1 , and the vectors x2 and v1 into the formula for v2 .
x2 • v1
v2 = x2 − v
v1 • v1 1
5 −1
−6 − 30 3
v2 = −
−1 15 2
−5 1
5 −1 3
−6 3 0
+2 =
−1 2 3
−5 1 −3
3
0
This means that v2 = .
3
−3
7 −1 3
4 3 0
Use x3 = , v1 = , and v2 = to find v3 using the formula shown below.
7 2 3
−4 1 −3
x3 • v1 x3 • v2
v3 = x3 − v1 − v
v1 • v1 v2 • v2 2
Compute the inner products that appear in this formula. First compute x3 • v1 .
x3 • v1 = (7)( − 1) + (4)(3) + (7)(2) + ( − 4)(1)

x3 • v1 = 15
Recall that v1 • v1 = 15.
Compute x3 • v2 .
x3 • v2 = (7)(3) + (4)(0) + (7)(3) + ( − 4)( − 3)

x3 • v2 = 54
Compute v2 • v2 .
v2 • v2 = (3)(3) + (0)(0) + (3)(3) + ( − 3)( − 3)

v2 • v2 = 27
Substitute 15 for x3 • v1 , 15 for v1 • v1 , 54 for x3 • v2 , 27 for v2 • v2 , and the vectors x3 , v1 , and v2 into the formula for v3 .
x3 • v1 x3 • v2
v3 = x3 − v1 − v
v1 • v1 v2 • v2 2
7 −1 3
4 15 3 54 0
v3 = − −
7 15 2 27 3
−4 1 −3
7 −1 3 2
4 3 0 1
− −2 =
7 2 3 −1
−4 1 −3 1
2
1
This means that v3 = , and so an orthogonal basis for the column space of the given matrix is
−1
1
−1 3 2
3 0 1
, , .
2 3 −1
1 −3 1

Linear Alg
Find an orthogonal basis for the column space of the matrix to the right. 1 4 6
−1 −2 1
0 4 4
1 6 3
1 4 8
Given a basis {x1 ,...,xp } for a nonzero subspace W of ℝn , define v1 ,...,vp as follows.
v1 = x1
x2 • v1
v2 = x2 − v
v1 • v1 1
x3 • v1 x3 • v2
v3 = x3 − v1 − v
v1 • v1 v2 • v2 2
⋯
xp • v1 xp • v2 xp • vp − 1
vp = xp − v1 − v2 − ⋯ − v
v1 • v1 v2 • v2 vp − 1 • vp − 1 p − 1
Then {v1 ,...,vp } is an orthogonal basis for W. This is called theGram-Schmidt process.
1 4 6
−1 −2 1
Let x1 = 0 , x2 = 4 , and x3 = 4 . Notice that these vectors are linearly independent and hence form a basis for a
1 6 3
1 4 8
nonzero subspace W of ℝn .
Since the vectors above form a basis for a nonzerosubspace, theGram-Schmidt process can be applied. Let v1 = x1 , so that
1
−1
v1 = 0 . Use this and x2 to find v2 .
1
1
x2 • v1
Recall that v2 = x2 − v . Compute the inner products that appear in this formula. First compute x2 • v1 .
v1 • v1 1
x2 • v1 = (4)(1) + ( − 2)( − 1) + (4)(0) + (6)(1) + (4)(1)

x2 • v1 = 16
Next compute v1 • v1 .
v1 • v1 = (1)(1) + ( − 1)( − 1) + (0)(0) + (1)(1) + (1)(1)
v1 • v1 = 4
Substitute 16 for x2 • v1 , 4 for v1 • v1 , and the vectors x2 and v1 into the formula for v2 .
x2 • v1
v2 = x2 − v
v1 • v1 1
4 1
−2 −1
16
v2 = 4 − 0
4
6 1
4 1
4 1 0
−2 −1 2
4 −4 0 = 4
6 1 2
4 1 0
0
2
This means that v2 = 4 .
2
0
6 1 0
1 −1 2
Use x3 = 4 , v1 = 0 , and v2 = 4 to find v3 using the formula shown below.
3 1 2
8 1 0
x3 • v1 x3 • v2
v3 = x3 − v1 − v
v1 • v1 v2 • v2 2
Compute the inner products that appear in this formula. First compute x3 • v1 .
x3 • v1 = (6)(1) + (1)( − 1) + (4)(0) + (3)(1) + (8)(1)

x3 • v1 = 16
Recall that v1 • v1 = 4.
Compute x3 • v2 .
x3 • v2 = (6)(0) + (1)(2) + (4)(4) + (3)(2) + (8)(0)
x3 • v2 = 24
Compute v2 • v2 .
v2 • v2 = (0)(0) + (2)(2) + (4)(4) + (2)(2) + (0)(0)

v2 • v2 = 24
Substitute 16 for x3 • v1 , 4 for v1 • v1 , 24 for x3 • v2 , 24 for v2 • v2 , and the vectors x3 , v1 , and v2 into the formula for v3 .
x3 • v1 x3 • v2
v3 = x3 − v1 − v
v1 • v1 v2 • v2 2
6 1 0
1 −1 2
16 24
v3 = 4 − 0 − 4
4 24
3 1 2
8 1 0
6 1 0 2
1 −1 2 3
4 −4 0 − 4 = 0
3 1 2 −3
8 1 0 4
2 1 0 2
3 −1 2 3
This means that v3 = 0 , and so an orthogonal basis for the column space of the given matrix is 0 , 4 , 0 .
−3 1 2 −3
4 1 0 4

Student: Kyle Hoang Date: 04/15/22 Instructor: Larissa Williamson Course: MAS3114-Assignment: M30 (HW) Approximation &

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Student: Kyle Hoang Date: 04/15/22 Instructor: Larissa Williamson Course: MAS3114-Assignment: M30 (HW) Approximation &

Uploaded by

Copyright:

Available Formats

4/15/22, 1:20 PM M30(HW) Approximation & Gram-Schmidt Process (6.

Instructor: Larissa Williamson

Find the closest point to y in the subspace W spanned by v1 and v2 .

and z = y − y . The vector y is called the orthogonal projection of y onto W.

Substitute the components of v1 and v2 into the expression for v1 • v2 .

v1 • v2 = (1)(6) + (0)(1) + ( − 1)(0) + ( − 2)(3)

v1 • v2 = (1)(6) + (0)(1) + ( − 1)(0) + ( − 2)(3)

Substitute for v1 and v2 and simplify the resulting expression.

Instructor: Larissa Williamson

Find the best approximation to z by vectors of the form c1 v1 + c2 v2 .

Compute the appropriate inner products. Find v1 • v2 .

Multiply and add to find z .

Thus, by the Best Approximation​Theorem, the best approximation of z is as follows.

Instructor: Larissa Williamson

u1 , ..., up is any orthogonal basis of​W, then

and z = y − y . The vector y is called the orthogonal projection of y onto W.

Use this formula to find y . Begin by finding y • u1 .

Finally, simplify the expression.

Instructor: Larissa Williamson

Instructor: Larissa Williamson

a. Let U be the 2 × 1 matrix whose only column is u1 . Compute UT U and UUT .

Instructor: Larissa Williamson

To find u3 ​, first compute the inner products. Compute u3 • u1 .

u3 • u1 = ​(0​)( − 1​) + ​(0​)( − 1​) + ​(1​)(1​)

u1 • u1 = ​( − 1​)( − 1​) + ​( − 1​)( − 1​) + ​(1​)(1​)

u3 • u2 = ​(0​)(1​) + ​(0​)( − 2​) + ​(1​)( − 1​)

u2 • u2 = ​(1​)(1​) + ​( − 2​)( − 2​) + ​( − 1)( − 1​)

Simplify this result.

Substitute u3 and u3 into the formula for v.

Simplify this result.

Instructor: Larissa Williamson

C. The statement is true​because, since z is orthogonal to u1 and u2 ​, it is orthogonal to every

B. The statement is false because y − projW y is in W and so cannot be orthogonal to W.

Instructor: Larissa Williamson

a. If W is a subspace of ℝn and if v is in both W and W ​, then v must be the zero vector.

projection must be onto the subspace spanned by

c. If y = z1 + z2 ​, where z1 is in a subspace W of ℝn and z2 is in W ​, then z1 must be the orthogonal projection of y onto W.

d. The best approximation to y by elements of a subspace of W is given by the vector y − projW y.

Instructor: Larissa Williamson

The first step in the​Gram-Schmidt process is to define v1 = x1 .

Complete the calculation of v2 .

Verify this basis is orthogonal by showing that v1 • v2 = 0.

(6 • − ​3) + ​( − 6 • − ​4) + ​(6 • − ​1) = 0

Instructor: Larissa Williamson

5 5 When given the orthogonal​basis, v1 , v2 ​, the orthonormal

Begin by finding v1 to normalize v1 .

Now find v2 to normalize v2 .

The orthonormal basis of the subspace spanned by the

Instructor: Larissa Williamson

x2 • v1 = ​(5​)( − 1​) + ​( − 6​)(3​) + ​( − 1​)(2​) + ​( − 5​)(1​)

v1 • v1 = ​( − 1​)( − 1​) + ​(3​)(3​) + ​(2​)(2​) + ​(1​)(1​)

Simplify this result.

x3 • v1 = ​(7​)( − 1​) + ​(4​)(3​) + ​(7)(2​) + ( − 4​)(1​)

Recall that v1 • v1 = 15.

x3 • v2 = ​(7​)(3​) + ​(4​)(0​) + ​(7​)(3​) + ​( − 4​)( − 3​)

v2 • v2 = ​(3​)(3​) + ​(0​)(0​) + ​(3​)(3​) + ​( − 3​)( − 3​)

Simplify this result.

Instructor: Larissa Williamson

x2 • v1 = ​(4​)(1​) + ​( − 2​)( − 1​) + ​(4​)(0​) + ​(6​)(1​) + ​(4​)(1​)

Simplify this result.

x3 • v1 = ​(6​)(1​) + ​(1​)( − 1​) + ​(4)(0​) + (3​)(1​) + ​(8​)(1​)

v2 • v2 = ​(0​)(0​) + ​(2​)(2​) + ​(4​)(4​) + ​(2​)(2​) + ​(0​)(0​)

Simplify this result.

Thus, by the Best ApproximationTheorem, the best approximation of z is as follows.

u1 , ..., up is any orthogonal basis ofW, then

To find u3 , first compute the inner products. Compute u3 • u1 .

u3 • u1 = (0)( − 1) + (0)( − 1) + (1)(1)

u1 • u1 = ( − 1)( − 1) + ( − 1)( − 1) + (1)(1)

u3 • u2 = (0)(1) + (0)( − 2) + (1)( − 1)

u2 • u2 = (1)(1) + ( − 2)( − 2) + ( − 1)( − 1)

C. The statement is truebecause, since z is orthogonal to u1 and u2 , it is orthogonal to every

a. If W is a subspace of ℝn and if v is in both W and W , then v must be the zero vector.

c. If y = z1 + z2 , where z1 is in a subspace W of ℝn and z2 is in W , then z1 must be the orthogonal projection of y onto W.

The first step in theGram-Schmidt process is to define v1 = x1 .

(6 • − 3) + ( − 6 • − 4) + (6 • − 1) = 0

5 5 When given the orthogonalbasis, v1 , v2 , the orthonormal

x2 • v1 = (5)( − 1) + ( − 6)(3) + ( − 1)(2) + ( − 5)(1)

v1 • v1 = ( − 1)( − 1) + (3)(3) + (2)(2) + (1)(1)

x3 • v1 = (7)( − 1) + (4)(3) + (7)(2) + ( − 4)(1)

x3 • v2 = (7)(3) + (4)(0) + (7)(3) + ( − 4)( − 3)

v2 • v2 = (3)(3) + (0)(0) + (3)(3) + ( − 3)( − 3)

x2 • v1 = (4)(1) + ( − 2)( − 1) + (4)(0) + (6)(1) + (4)(1)

x3 • v1 = (6)(1) + (1)( − 1) + (4)(0) + (3)(1) + (8)(1)

v2 • v2 = (0)(0) + (2)(2) + (4)(4) + (2)(2) + (0)(0)