91Ó°ÊÓ

A vector norm is a function that assigns a positive length or size to each vector in a vector space, except the zero vector which is assigned a length of zero. It is denoted as \( \|x\| \) for a vector \( x \). There are different types of vector norms, each measuring the size of the vector in slightly different ways:

• **Euclidean Norm:** This is the most common type of vector norm, also known as the \( L^2 \) norm, which calculates the "straight-line" distance from the origin to the point \( x \). It is represented by \( \|x\|_2 = \sqrt{x_1^2 + x_2^2 + ... + x_n^2} \).
• **Manhattan Norm:** Also known as the \( L^1 \) norm, it is the sum of the absolute values of the vector components, \( \|x\|_1 = |x_1| + |x_2| + ... + |x_n| \).
• **Infinity Norm:** The maximum absolute value of the vector components, \( \|x\|_\infty = \max(|x_1|, |x_2|, ..., |x_n|) \).

Each norm gives a different way of measuring the vector's magnitude, and selecting which one to use depends on the context of the problem. A fundamental property is that for any vector \( x \), \( \|x\| \geq 0 \). Understanding vector norms is crucial because they form the building block for defining subordinate matrix norms.

The matrix-vector product is the result of multiplying a matrix with a vector, producing another vector. If \( A \) is an \( n \times n \) matrix and \( x \) is a vector with \( n \) elements, then the matrix-vector product \( Ax \) is defined in such a way that each element of the resulting vector is a linear combination of the columns of the matrix \( A \).

Mathematically, this can be expressed as:

• The vector \( Ax \) at position \( i \) is equal to the dot product of the \( i \)-th row of \( A \) and the vector \( x \), or \( (Ax)_i = a_{i1}x_1 + a_{i2}x_2 + ... + a_{in}x_n \).

This operation is fundamental in linear algebra and has applications in various fields such as computer graphics, machine learning, and engineering. The matrix-vector product is instrumental in expressing systems of linear equations and transformations.

In the context of subordinate matrix norms, we aim to find a specific vector that satisfies certain conditions involving matrix norms. The statement we are exploring asserts that, for any matrix \( A \) and its subordinate matrix norm \( \|A\| \), there exists a non-zero vector \( x \) such that \( \|Ax\| = \|A\|\|x\| \).

To prove this, consider that the subordinate matrix norm is defined as \( \|A\| = \max_{\|x\|=1} \|Ax\| \). This implies there exists at least one unit vector \( x \) such that \( \|Ax\| \) achieves this maximum value \( \|A\| \).

• The phrase "unit vector" means \( \|x\| = 1 \), so these unit vectors are not the zero vector, automatically fulfilling the non-zero criterion.
• This vector \( x \) shows that the matrix-vector product maintains the equality \( \|Ax\| = \|A\|\|x\| \). It reflects that the scalar stretching by the matrix \( A \) has a specific vector direction for which the matrix applies its maximum effect.

This understanding is pivotal because it bridges the gap between theoretical norm definitions and practical computation, linking matrix behavior to vector essence.