Gilbert Strang-Linear Algebra-Orthogonality
Orthognality
Orthogonality of the four spaces
Def.1 Orthognality:
Two subspaces and of a vector space are if every vector is perpendicular to every vector in .
Orthogonal subspaces
The surface of the floor and the wall is perpendicular, but these two spaces are not orthogonal, since we can still find two vectors in each space that are not perpendicular.
Null space and Row Space are orthogonal
If , then we have .
Inner product of and the matrix equals . Similiarly we can prove that Row space and left null space are orthogonal.
Orthognal Complemen
Row space and Null space split into two orthogonal subspaces. For example for matrix , , and is its basis. Its Null space span a 2 dimensional surface that is perpendicular to the vector
Def.2 Orthogonal Complement:
The Orthogonal complement of a subspace contains every vector that is perpendicular to ,whichis denoted by .
Important:
is the orthogonal complement of the Row Space .(in )
is the orthogonal complement of the Comlumn Space (in )
Projection
Question: How to solve and un-solvable formula
When the matrix is rectangular, , as there are multitude number on the left side of the formula, there is a large likely that some “bad” data are mixed in, making the formula unsolvable. What linear algebra does is to find the “best solution” in such circumstance. The matrix plays an important role.
where is invertible. However it is not always invertible. Invertibility requires row vectors of are linear independent. The core of this chapter is to solve to find the best solution.
E.g. find solution of where are independent.
The best solution in this case is , where is the projection of on . Since , then we can get the equation that:
Projection Matrix
Let us describe the problem in projection matrix, which is , where is the projection matrix. Since
Therefore, when is unsolvable, we can
Firstly, project onto using projection matrix and get ;
Secondly, solve to find the best solution.
High dimentional projection
In
If row vector make a set of base, then the surface is the row space of the matrix . It is known that the vector is in the surface, then we have:
Least Square Approximation
Find the best solution for an equation using Projection Matrix. For example,if we want fit 3 point with a straight line:
3 points:
Curve equationL
In the matrix form we have :
Orthonormal
Def. Orthonormal: if the colomn vectors of a matrix are unit vector that are perpendicular to each other, then they(the vectors) are linear independent and orthonormal.
Def. Orthogonal Matrix if the column vectors of a square matrix are orthonormal, then the square matrix itself is orthonormal matrix.
Feature: the inner product of orthonormal with itsel is identity matrix. Say is orthonormal matrix
*Note:this feature also applies to the situation where the is not a square
References
- Gilbert, Strang. Introduction to Linear Algebra, 4th Edition[M]. MA 02482 US:Wellsley-Cambridge Press, 2009.
- ls317842927. 线性代数学习笔记4 [EB/OL]. http://blog.****.net/ls317842927/article/details/53309746.
- MIT-线性代数笔记(上)[EB/OL]. http://www.docin.com/p-1489481927.html.