Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
2 views

Square Matrices

The document discusses concepts related to square matrices, including eigenvalues and eigenvectors, matrix inverses, and pseudoinverses. It explains the conditions under which eigenvalues and eigenvectors exist, as well as properties of the pseudoinverse for singular matrices. Additionally, it covers symmetric matrices and basic matrix identities, and introduces random variables in the context of probability.

Uploaded by

Paulo Cabeceira
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
2 views

Square Matrices

The document discusses concepts related to square matrices, including eigenvalues and eigenvectors, matrix inverses, and pseudoinverses. It explains the conditions under which eigenvalues and eigenvectors exist, as well as properties of the pseudoinverse for singular matrices. Additionally, it covers symmetric matrices and basic matrix identities, and introduces random variables in the context of probability.

Uploaded by

Paulo Cabeceira
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 9

Square Matrices

If we consider matrix A as a mapping from R n to R n by stretching,


squeezing, rotating, etc, and for some nonzero vectors v , the mapping may
be particularly simple: Av is just a scaled version of v , Av = λv .
For any λ and v satisfying Av = λv , v is called an eigenvector of A and
λ is the corresponding eigenvalue.

Av = λv
Av − λv = 0
Av − λIv = 0
, where I is the identity matrix

(A − λI)v = 0

If v non-zero, the equation will have a solution only if |A − λI| = 0


(©2022 Tapomayukh Bhattacharjee) Foundations of Robotics August 22, 2022 15 / 43
Example
Eigenvalue Example:
2 1
A=
1 2
2−λ 1
|A − λI| = = λ2 − 4λ + 3 = (λ − 3)(λ − 1) = 0
1 2−λ
   
1 1 1
When λ = 1, (A − λI)v = 0, v = 0, v = . When λ = 3,
1 1 −1
   
−1 1 1
(A − λI)v = 0, v = 0, v = .
1 −1 1
 
1
eigenvector : v1 = , λ1 = 1
−1
 
1
eigenvector : v2 = , λ2 = 3
1

(©2022 Tapomayukh Bhattacharjee) Foundations of Robotics August 22, 2022 16 / 43


Inverse

For matrix A ∈ Rm×n ,


the matrix inverse, A−1 is the unique matrix that satisfies
AA−1 = A−1 A = I .
Matrix inverse can be used to solve Ax = b, where x = A−1 b.

(©2022 Tapomayukh Bhattacharjee) Foundations of Robotics August 22, 2022 17 / 43


Pseudoinverse

If A is not full rank (it is singular), then the inverse does not exist, but we
can still calculate the Moore-Penrose pseudoinverse of A, denoted A† . The
peudoinverse has ”inverse-like” properties and can be used to find
solutions or approximate solutions to Ax = b, where x = A† b. The
pseudoinverse A† is equivalent to inverse A−1 when A is invertible.

A† A = A

(©2022 Tapomayukh Bhattacharjee) Foundations of Robotics August 22, 2022 18 / 43


Pseudoinverse Properties

For A† ∈ Rmxn of any real matrix A ∈ Rnxm :

AA† A = A

A† AA† = A†
AA† is symmetric
A† A is symmetric

(©2022 Tapomayukh Bhattacharjee) Foundations of Robotics August 22, 2022 19 / 43


Symmetric Matrices

A square matrix A is symmetric if it is equal to its transpose, A = AT . A


matrix A is skew symmetric if A = −AT .

(©2022 Tapomayukh Bhattacharjee) Foundations of Robotics August 22, 2022 20 / 43


Basic Matrix Identities

(AB)C = A(BC )
(AT )−1 = (A−1 )T
(Ax)T = x T AT
(AB)T = B T AT
(ABC ...)T = ...C T B T AT
(AB)−1 = B −1 A−1
(ABC ...)−1 = ...C −1 B −1 A−1

(©2022 Tapomayukh Bhattacharjee) Foundations of Robotics August 22, 2022 21 / 43


Topics

Part 1: Linear Algebra


Part 2: Probability

Content Courtesy: Mathematical Statistics and Data Analysis, 3rd Edition


(©2022 Tapomayukh Bhattacharjee) Foundations of Robotics August 22, 2022 22 / 43
Random Variables

A random variable is a function from sample space Ω to real numbers.


Example: A coin is tossed 3 times, and we can observe the sequence of
heads and tails:

Ω = {hhh, hht, htt, hth, ttt, tth, thh, tht}

The total number of heads is a random variable defined on Ω, so are the


total number of tails and the number of heads minus the number of tails.
In general, we denote random variables with uppercase letters. We will use
lowercase letters to denote the values that random variables can take on.
For example, the possible values of random variable X are x1 .x2 , ...

(©2022 Tapomayukh Bhattacharjee) Foundations of Robotics August 22, 2022 23 / 43

You might also like