The Kernel (Null Space) of a Linear Transformation
What you'll learn: You'll discover which vectors get "collapsed" to zero by a linear transformation and what that tells us about whether the transformation is one-to-one.
Understanding the Kernel
The kernel (or null space) of a linear transformation T: V → W is the set of all vectors in V that map to the zero vector in W. In symbols:
ker(T) = {v ∈ V : T(v) = 0}
Think of it as the "input that produces nothing." These are the vectors that the transformation completely "crushes" down to zero.
Why It Matters: Injectivity
The kernel reveals whether your transformation is injective (one-to-one):
- If ker(T) = {0} (only the zero vector maps to zero), then T is injective. Different inputs always produce different outputs.
- If ker(T) contains non-zero vectors, then T is not injective. Multiple inputs can produce the same output.
An Analogy
Imagine a transformation as a photo filter. If the kernel contains only the "blank image," then every distinct photo produces a distinct filtered result. But if some non-blank photos also turn completely black (zero), the filter isn't one-to-one—information is lost.
Finding the Kernel
To find ker(T), solve the equation T(v) = 0. This is exactly like finding the null space of a matrix: set up the homogeneous system and find all solutions.
For example, if T is represented by matrix A, then ker(T) is the null space of A—all vectors x where Ax = 0.
The kernel is always a subspace of the domain V (it contains 0, and is closed under addition and scalar multiplication).
Key Takeaway: The kernel of a linear transformation is the set of all vectors mapping to zero; a transformation is injective if and only if its kernel contains only the zero vector.