0% found this document useful (0 votes)
10 views

1 Vectors

Vectors can be represented as ordered lists of numbers and are used throughout linear algebra and machine learning to represent various types of data. Some key examples of vectors include location vectors in 2D or 3D space, images represented as vectors of pixel intensities, word count vectors for text analysis, and feature vectors containing attributes of data points. Vectors can be manipulated through operations like addition, subtraction, and scalar multiplication according to simple rules.

Uploaded by

Xiya Luo
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
10 views

1 Vectors

Vectors can be represented as ordered lists of numbers and are used throughout linear algebra and machine learning to represent various types of data. Some key examples of vectors include location vectors in 2D or 3D space, images represented as vectors of pixel intensities, word count vectors for text analysis, and feature vectors containing attributes of data points. Vectors can be manipulated through operations like addition, subtraction, and scalar multiplication according to simple rules.

Uploaded by

Xiya Luo
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 29

1: Vectors

Vectors

Examples

Addition and scalar multiplication

Inner product

Complexity

Source: Introduction to Applied Linear Algebra, Boyd & Vandenberghe

D. Matsypura, A. Wong, QBUS1040 1/29


Vectors 1

• a vector is an ordered list of numbers


• we’ll use two styles of notation:
! $
−1.2
" 0.0%
" %
# 3.6& or (−1.2, 0.0, 3.6, −7.2)
−7.2
• numbers in a vector are called entries, coefficients, or elements
• length of vector is its size, length, or dimension
• vector above has dimension 4; its third entry is 3.6
• vector of length n is called an n-vector
• numbers are called scalars

1: Vectors 2/29
Vectors via symbols 1

• we’ll use symbols to denote vectors, e.g., a, X, p, β, E aut

• other conventions (that you should not use in this course): g, "a

• ith element of n-vector a is denoted ai

• if a = (5, 7, 1, 4, 6) then, a3 = 1

• in ai , i is the index

• for an n-vector, indices run from i = 1 to i = n

• ar:s = (ar , . . . , as ) is a subvector of a, and a2:4 = (7, 1, 4)

• warning: sometimes ai refers to the ith vector in a list of vectors

• two vectors a and b of the same size are equal if ai = bi for all i

• we overload “=” and write this as a = b

1: Vectors 3/29
Block vectors 2

• suppose b, c, and d are vectors with sizes m, n, p

• the stacked vector or concatenation (of b, c, and d) is


! $
b
a = # c & = (b, c, d)
d

• also called a block vector, with (block) entries b, c, d

• a has size m + n + p

a = (b1 , b2 , . . . , bm , c1 , c2 , . . . , cn , d1 , d2 , . . . , dp )

1: Vectors 4/29
Zero, ones, and unit vectors 3,4

• n-vector with all entries 0 is denoted 0n or just 0

• n-vector with all entries 1 is denoted 1n or just 1

• a unit vector has one entry 1 and all others 0

• denoted ei where i is entry that is 1

• unit vectors of length 3:


! $ ! $ ! $
1 0 0
e1 = # 0 & , e2 = # 1 & , e3 = # 0 &
0 0 1

1: Vectors 5/29
Sparsity 5

• a vector is sparse if many of its entries are 0

• can be stored and manipulated efficiently on a computer

• nnz(x) is number of entries that are nonzero

• examples: zero vectors, unit vectors

1: Vectors 6/29
Outline

Vectors

Examples

Addition and scalar multiplication

Inner product

Complexity

1: Vectors 7/29
Location or displacement in 2-D or 3-D 6

• 2-vector (x1 , x2 ) can represent a location or a displacement in 2-D

1: Vectors 8/29
Images, video

Monochrome (black and white) image


grayscale values of M × N pixels stored as M N -vector (e.g., row-wise)
5
5
0
0
0 .6
0 .0
0 .2
0 .2 ! $ ! $
x1 0.65
" x2 % "0.05%
" % " %
" x3 % "0.20%
" % " %
" . % " . %
x = " .. % = " .. %
" % " %
"x62 % "0.28%
" % " %
#x63 & #0.00&
x64 0.90
0
8
0
0
1 .0
0 .2
0 .0
0 .9

Color image: 3M N -vectors with R, G, B values of the M N pixels

Video: KM N -vector represents K monochrome images of M × N pixels

1: Vectors 9/29
Portfolio vector

Portfolio
• n-vector represents stock portfolio or investment in n assets

• ith element is amount invested in asset i

• elements can be number of shares, dollar values, or fractions of total dollar


amount

p = (5000, 2000, 8000, 3000)

1: Vectors 10/29
Resource vector

Resource vector
• elements of n-vector represent quantities of n resources or commodities

• sign indicates whether quantity is held or owed, produced or consumed, ...

• example: bill of materials gives quantities needed to create a product

r = ( −2, 1, 0, −1)

1: Vectors 11/29
Feature vectors

contain values of variables or attributes that describe members of a set.

Examples
• age, weight, blood pressure, gender, . . . of patients

• size, number of bedrooms, list price, . . . of houses in an inventory

Area (m2 ) Bedrooms Bathrooms Cars Pets Weekly rental price ($)
252 3 2 2 1 1400

Note
• vector elements can represent very different quantities, in different units

• can contain categorical features (e.g., 0/1 for male/female)

• ordering has no particular meaning

1: Vectors 12/29
Word count vectors 7

• a short document:

Word count vectors are used in computer based document analysis.


Each entry of the word count vector is the number of times the
associated dictionary word appears in the document.

• a small dictionary (left) and word count vector (right)


! $
word 3
in "2%
" %
number "1%
" %
horse "0%
" %
the #4&
document 2

• dictionaries used in practice are much larger

1: Vectors 13/29
Polynomials 8

a polynomial of degree n − 1

f (t) = c1 + c2 t + c3 t2 + · · · + cn tn−1

can be represented by an n-vector (c1 , c2 , . . . , cn )

Example
f (t) = 3 + t − 5t2 + 2t3
would be represented by the n-vector (3, 1, −5, 2)

1: Vectors 14/29
Outline

Vectors

Examples

Addition and scalar multiplication

Inner product

Complexity

1: Vectors 15/29
Vector addition 9

• n-vectors a and b can be added, with sum denoted a + b

• to get the sum, add corresponding entries:


! $ ! $ ! $
0 1 1
# 7 &+# 2 &=# 9 &
3 0 3

• subtraction is similar

Properties of vector addition:


• commutative: a + b = b + a

• associative: (a + b) + c = a + (b + c)
(so we can write both as a + b + c)

• a+0=0+a=a

• a−a=0
1: Vectors 16/29
Scalar-vector multiplication 10

• scalar β and n-vector a can be multiplied

βa = (βa1 , . . . , βan )
• also denoted aβ
• example:
! $ ! $
1 −2
(−2) # 9 & = # −18 &
6 −12

Properties of scalar-vector multiplication:


• associative: (βγ)a = β(γa)

• left distributive: (β + γ)a = βa + γa

• right distributive: β(a + b) = βa + βb

1: Vectors 17/29
Linear combinations 11

• for vectors a1 , . . . , am and scalars β1 , . . . , βm ,

β1 a 1 + · · · + βm a m

is a linear combination of the vectors

• β1 , . . . , βm are the coefficients

• a very important concept

• a simple identity: for any n-vector b,

b = b1 e 1 + · · · + bn e n

1: Vectors 18/29
Outline

Vectors

Examples

Addition and scalar multiplication

Inner product

Complexity

1: Vectors 19/29
Inner product 12

• inner product (or dot product) of n-vectors a and b is

aT b = a1 b1 + a2 b2 + · · · + an bn

• other notation used (that you should not use in this course):
〈a, b〉, 〈a|b〉, (a, b), a · b

• example:
! $T ! $
−1 1
# 2 & # 0 & = (−1)(1) + (2)(0) + (2)(−3) = −7
2 −3

1: Vectors 20/29
Properties of inner product 12

• aT a = a21 + a22 + · · · + a2n ≥ 0

• aT a = 0 if and only if a = 0

• commutative: aT b = bT a

• associative with scalar multiplication: (γa)T b = γ(aT b)

• distributive with vector addition: (a + b)T c = aT c + bT c

can combine these to get, for example,

(a + b)T (c + d) = aT c + aT d + bT c + bT d

1: Vectors 21/29
Simple examples

• eTi a = ai (inner product with ith unit vector picks out ith entry)

• (ei − ej )T a = ai − aj (differencing)

• 1 T a = a1 + · · · + an (sum of entries)

• (1/n)T a = (a1 + · · · + an )/n (average of entries)

1: Vectors 22/29
Example: Weighted sum

• w is vector of weights

• f is vector of features

• inner product wT f = w1 f1 + w2 f2 + · · · + wn fn is total score

Example

w = ( 0.25, 0.25, 0.5) (assessment weights)


f =( 60, 80, 70) (assessment marks)

wT f = 0.25 × 60 + 0.25 × 80 + 0.5 × 70


= 15 + 20 + 35
= 70

1: Vectors 23/29
Example: Cost

• p is vector of prices of n goods

• q is vector of quantities purchased

• inner product pT q = p1 q1 + p2 q2 + · · · + pn qn is total cost


Example

p = ( 1.00, 2.50, 5.00) (fruit prices)


q=( 2, 4, 1) (fruit quantity)

1: Vectors 24/29
Example: Portfolio return

• h is portfolio vector, with hi the dollar value of asset i held

• r is vector of fractional returns over the investment period:

pfinal
i − pinit
i
ri = , i = 1, . . . , n
pinit
i

where pinit
i and pfinal
i are the prices of asset i at the beginning and end of
the period

• rT h = r1 h1 + · · · + rn hn is the total return, in dollars, over the period

1: Vectors 25/29
Example: Portfolio return
Example: We have two companies we have invested in. We invested $200 into
company 1 and $100 into company 2. This means that:

h = (200, 100)

pfinal
1 − pinit
1 27.10 − 54.20
r1 = init
= = −0.5
p1 54.20
pfinal
2 − pinit
2 50.00 − 40.00
r2 = = = 0.25
pinit
2 40.00
Our fractional returns are then:
r = (−0.5, 0.25)
Our total returns are then:
rT h = −0.5 × 200 + 0.25 × 100 = −100 + 25 = −75
1: Vectors 26/29
Outline

Vectors

Examples

Addition and scalar multiplication

Inner product

Complexity

1: Vectors 27/29
Flop counts

• computers store (real) numbers in floating point format

• basic arithmetic operations (+, −, ∗, /, . . .) are called floating point


operations or flops

• complexity of an algorithm or operation: total number of flops needed, as


a function of the input dimension(s)

• crude approximation of time to execute:

number of operations (flops)


run time ≈
computer speed (flops per second)

• current computers are around 1Gflop/sec (109 flops/sec)

1: Vectors 28/29
Complexity

Operation count (flop count)


• total number of operations in an algorithm
• in linear algebra, typically, a polynomial of the dimensions of the problem

Dominant term: the highest-order term in the flop count


1 3 1
n + 100n2 + 10n + 5 ≈ n3
3 3
Order: the power in the dominant term
1 3
n + 10n2 + 500 = order n3
3

Examples of vector operations in this lecture (for vectors of size n):


• addition, substraction: n flops
• scalar multiplication: n flops
• inner product: 2n − 1 ≈ 2n flops
these operations are all of order n
1: Vectors 29/29

You might also like