Final Study Cards Flashcards

1
Q

What is a scalar valued function?

A

A scalar valued function is a function which maps R^n onto R

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

When is a scalar valued function linear?

A

A scalar valued function is linear if it has the form f(x1,…,xn) = a1x1 + a2x2 + ··· + anxn
i.e., it is affine with b = 0, or equivalently with f(0) = 0.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

When is a scalar valued function affline?

A

A scalar valued function is affline if it has the form f(x1,…,xn) = a1x1 + a2x2 + ···+ anxn + b for some numbers a1,…,an,b (so b = f(0)).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What is the derivative matrix of a vector valued function?

A

The derivative matrix of a vector valued function f(x) = {f_1(x), f_2(x), f_3(x),…,f_n(x)
is the matrix such that
| f_1/dx_1 f_1/dx_2 f_1/dx_3 … f_1/dx_n|
| f_2/dx_1 f_2/dx_2 f_2/dx_3 … f_2/dx_n|
| f_3/dx_1 f_3/dx_2 f_3/dx_3 … f_3/dx_n|
| …
| f_n/dx_1 f_n/dx_2 f_n/dx_3 … f_n/dx_n|

f_1/dx_1 f_1/dx_2 f_1/dx_3 … f1/dx_n|

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What are the two linear approximations for a function F at x using the derivative matrix (where a is a close value)?

A

f(x) ≈f(a) + ((Df)(a))(x −a)
and
f(a + h) ≈f(a) + ((Df)(a)) h

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What is the linearity principle?

A

for c1,c2 ∈R and v1,v2 ∈R^2 we have f(c1v1 + c2v2) = c1f(v1) + c2f(v2).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

When is a function g linear?

A

When g(cx) = cg(x), g(x + y) = g(x) + g(y)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What is the rotation matrix for R^2?

A

Aθ =
|cos θ −sin θ|
|sin θ cos θ|
.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What is the identity matrix and what are it’s properties?

A

The identity matrix is an n x n square matrix such that its entries are equal to
|1 0 0 … 0|
|0 1 0 … 0|
|0 0 1 … 0|
|… |
|0 0 0 … 1|

The special property of the identity matrix is that any m x n matrix A multiplied by its respective n x n identity matrix is equal to A

so A*I_n = A

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

What are the important properties of matrix multiplication

A

(MM1) It recovers matrix-vector multiplication: if A is an m ×n matrix, and x ∈ R_n is thought of as an n ×1 matrix, the matrix-matrix product Ax is the same as the matrix-vector product.
(MM2) A(B + C) = AB + AC and (A′ + B′)C′ = A′C′ + B′C′. (These “distributive laws” are the
reason we call it matrix multiplication.)
(MM3) A(BC) = (AB)C, and A(cB) = (cA)B = c(AB) for any scalar c. In particular, taking C
to be an m ×1 matrix that is a column vector v by another name, A(Bv) = (AB)v.
(MM4) If A is an m ×n matrix, then I_mA = A = AI_n, where I_m is the m × m identity matrix and I_n is the n ×n identity matrix.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What are some important assertions about matrix multiplication

A

AB != BA
AB = AC does not imply B = C

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

How does one set up and solve a markov chain problem?

A

First create a matrix M which models the total of each team after each cycle. For the n_th step in the cycle, the number at each position equals to M^n-1th * the original values

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What is the multivariable chain rule at a point = (v1,…,vn) ∈ R^n?

A

(D(f ◦ g))(v) = (Df)(g(v)) (Dg)(v)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What is the definition of an inverse matrix B to a matrix A?

A

The inverse matrix B is a matrix such that BA = I_n and AB = I_n

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

If A is invertible, what is Ax = b equal to?

A

x = (A^-1)(b)
It is important to note that the position of the A is maintained relative to the vector it is multiplying (so x != (b)(A^-1))

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What is true of n x n matrices with respect to their inverse?

A

If A and B are n × n matrices that satisfy AB = In then A is invertible and B is its
inverse; i.e., automatically the other equation BA = In holds.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

How can you immediately tell if a set of vectors is not linearly independent?

A

If it has a vector which is not non-zero

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

When is a vector non-zero

A

When it has at least one entry which is not zero

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

The upper left to lower right diagonal of the inverse of a square matrix is always what

A

1/The original value of the position in the diagonal (aka the reciprocal)

20
Q

What is the determinant of an upper or lower triangular matrix?

A

The product of its diagonal entries

21
Q

When is a 2x2 matrix inversible?

A

When it’s determinant is not equal to 0

22
Q

What is the determinant of a 2x2 matrix

A

The determinant is ad-bc where the slots are assigned
| a b |
| c d |

23
Q

When is a matrix A invertible?

A

A is invertible precisely when Ax = 0 has x = 0 as its only solution, and A is non-invertible precisely when Ax = 0 has a nonzero solution.

24
Q

What are some important matrix inversion rules?

A

(i) When A is invertible you can “cancel A” by multiplying both sides by A−1 (but there is a
caveat; see the Warning below):
– Cancelling an invertible matrix on the left: if AB = AC and A is invertible then
B = C. This holds because you multiply both sides on the left by A−1.
– Cancelling an invertible matrix on the right: if BA = CA and A is invertible then
B = C. For this you need to multiply both sides on the right by A−1.
– Warning: our caveat is that if AB = CA, then you cannot cancel A on the left on one
side and on the right on the other, so you cannot conclude in this case that B = C, even
when A is invertible (see Example 18.4.1 below).
(ii) If A and B are both invertible n ×n matrices then AB is also invertible, and (AB)−1 =
B−1A−1 (note the switch of order of multiplication on the right side!); see Example
18.4.3 below for an illustration.

25
Q

What is the inverse matrix of a 2x2 matrix (if it is invertible)?

A

It’s inverse matrix is

(1/ad-bc)|d -b|
|-c a|
where the original matrix is
|a b|
|c d|

26
Q

Newton Whatever Whatever Convergence Thing

A

Add It Maybe

27
Q

When is a collection of vectors considered linearly independent?

A

When none of the vectors belong to the span of the others

28
Q

When are a collection of vectors linearly independent

A

A collection of vectors v_1,…,v_k ∈R_n is linearly independent precisely when the
only collection of scalars a_1,…,a_k for which
a_1v_1 + a_2v_2 + ···+ a_kv_k = 0
is a_1 = 0,a_2 = 0,…,a_k = 0.

29
Q

How do you do the Gramm-Schmidt process for a collection of vectors v_1, v_2, …, v_n

A

Let v1,…,vk be nonzero n-vectors with span V in Rn.
Let w1 = v1 and define B1 to be {w1}(an orthogonal basis for V1!).
Let w2 = v2 −Projw1(v2)
Let w3 = v3 −Projw_2(v3) - Projw_2(v3)
and so on, defining at the jth step wj = vj − ProjVj−1(vj)

30
Q

How do you determine if a set of vectors {v1,v2,…,vk} are linearly independent?

A

If dim(span({v1,v2,v3}) = k

31
Q

What is the dimension of the orthogonal compliment to linear subspace v∈R_n?

A

The dimension of the orthogonal compliment is n - dim(V)

32
Q

What is the transpose of a matrix?

A

It is the resulting vector when you flip the values of the matrix across the upper left to lower right diagonal of the original matrix

33
Q

What does the transpose allow you to do?

A

The transpose allows you to move a matrix across a dot product, so for example

(Ax) dot y = x dot (A_Transpose(y))

34
Q

What are the important properties of matrix algebra

A

A(B + C) = AB + AC, (A + B)C = AC + BC, A(BC) = (AB)C, and
AIn = A = ImA for an m ×n matrix A. But AB 6= BA in general!
* (From before) Sometimes matrices are invertible. When they are invertible, you can multiply
by their inverse to cancel them: for example, if AB = AC and A is invertible then B = C,
whereas if AB = CA (with invertible A) then you can’t conclude anything.
* (From before) Invertible matrices are always square; i.e., n×n for some n. Inversion reverses
the order of multiplication: (AB)−1 = B−1A−1 if A and B are both invertible.
* (New) The transpose of an m×n matrix is an n×m matrix, and transpose reverses the order
of multiplication: (AB)> = B>A>.
* (New) If A is invertible so is A>, and (A>)−1 = (A−1)>.
* (New) For v,w ∈Rn viewed as n×1 matrices, the 1 ×1 matrix product v>w equals [v·w].
This yields efficient manipulation of dot products of many vectors at once via matrix algebra.

35
Q

What are the properties of a symmetric matrix

A

A_Transpose = A and the matrix is square (i.e. n x n). The inverse of a symmetric matrix is always symetric

36
Q

When is a given matrix A orthogonal?

A

When its transpose times it is equal to the identity matrix ((A_transpose)*A = I_ n)

or

The n columns of A are an orthonormal collection of m-vectors

37
Q

What is the inverse of an orthogonal matrix?

A

Its transpose

38
Q

Given matrices A and B are orthogonal, what is true of their product?

A

Their product regardless of order will be orthogonal

39
Q

What is the column space of a matrix?

A

The span of its columns

40
Q

When is an upper or lower triangular matrix invertible?

A

When all its diagonal entries are non-zero

41
Q

How does one solve an equation given the LU decomposition?

A

Say that L(Ux) = b and then say that Ux = y = {y1,y2,…,yn} and solve for Ly = b using back substitution.
Then replug in the values of Ux into the equation to solve via forward substitution

42
Q

How does one solve an equation given the QR decomposition?

A

Since Q is orthogonal we can say that
QRx = b => Rx = (Q_Transpose)b
Solve this to get the answer

43
Q

How do you solve the inverse of a matrix given its LU decomposition?

A

The inverse of a matrix given it’s LU decomposition is equal to L^-1U^-1
To solve these state that L’ and U’ are the inverses with the recipricol values along the diagonal and the rest as unknown of L and U, and then say that L’L = I and U’U = I solving for the unknowns using backwards or forwards substitution

44
Q

How do you solve the inverse of a matrix given its QR decomposition?

A

You know how to do it you twat

45
Q

How to find the QR decomposition of a matrix?

A

Given a matrix of form A = {v1,v2,v3,…,vn} run gramm-schmidt on the given vectors to get w1,w2,w3,…,wn. Q = the orthonormal form of these vectors

To find R write v1,…,vn as their equivalent form in terms of w1,..,wn
Multiply the coefficients of the given vn equation by the magnitude of wn is equivalent to the column vn