Of Notes
Matrices
Table of Contents
A Matrix is some
Some nice properties
- Given
, its transpose is such that - If
then we call it a square matrix - If
then it's called symmetric - We call vectors matrices of one column
- If
for then is diagonal - If
we say it's an identity matrix - If you remove $i$th and $j$th row and column from
you get a minor
Intuition: A matrix fundamentally is a linear transformation operator which sometimes will "stretch" space in different ways.
Operations on Matrices
Sums
Very simple, we elementwise add the items:
Products
This is a bit more complicated because we take the sum of products of the rows of
Which we can simplify with sums:
Note how the first matrix fixes the row and iterates over the columns while the second matrix fixes the column and iterates over the rows.
Which expanded out for each term looks like:
This is only defined if
An important property is with transpose that
Matrix Determinant
This deserves a separate section.
Matrix Rank
Exponential of a Matrix
Special Cases - Structured Matrices
- Symmetric Matrix
- Hermitian Matrix
- Hessian Matrix
- Orthogonal
- Unitary
- Jacobian