Skip to content

Some properties

Let us look at some properties of the matrix \(X^TX\). First, if \(X\) is an \(m \times n\) matrix, then \(X^TX\) is an \(n \times n\) square matrix.

Symmetric

Property-1

\(X^TX\) is symmetric.

To see why:

\[ (X^TX)^T = X^T(X^T)^T=X^TX \]

Nullspace

Property-2

\(X\) and \(X^TX\) have the same nullspace.

To see why, let us take this in two steps:

If \(\theta \in N(X)\), then:

$$ \begin{aligned} X \theta &= 0\\ X^TX \theta &= 0\\ \end{aligned} $$ So, \(\theta \in N(X^TX\)). Going the other way, if \(\theta \in N(X^TX)\), then:

\[ \begin{aligned} X^TX \theta &= 0\\\\ \theta^T X^TX \theta &= 0\\\\ (X \theta)^T (X \theta) &= 0\\\\ X \theta &= 0 \end{aligned} \]

So, \(\theta \in N(X)\).

Invertibility

Property-3

If \(\text{rank}(X) = n\), then \(X^TX\) is invertible.

To see why, note that if \(\text{rank}(X) = n\), then the nullity of \(X\) is zero. Since \(X\) and \(X^TX\) have the same nullspace, nullity of \(X^TX\) is also zero. From this it follows that the rank of \(X^TX\) is \(n\). Since this is a full rank matrix, \(X^TX\) is invertible.