Principal Minors and Sylvester's Criterion
What You'll Learn
You'll learn how to test whether a symmetric matrix is positive definite by checking a sequence of determinants—no eigenvalue computation required.
The Core Idea
Recall that a symmetric matrix is positive definite if all its eigenvalues are positive. While computing eigenvalues works, it's often tedious. Sylvester's Criterion gives us a shortcut: just check the determinants of certain submatrices.
A leading principal minor of order k is the determinant of the top-left k×k submatrix. For an n×n matrix, you have n leading principal minors (one for each size from 1×1 up to n×n).
Sylvester's Criterion states:
A symmetric matrix is positive definite if and only if all its leading principal minors are positive.
Example
Consider the matrix:
A = [ 2 -1 ]
[-1 3 ]
- First leading principal minor (top-left 1×1): det([2]) = 2 > 0 ✓
- Second leading principal minor (full 2×2): det(A) = 2·3 − (−1)(−1) = 6 − 1 = 5 > 0 ✓
Both are positive, so A is positive definite.
Why It Works
Think of it as a "stability test" at each dimension. Each leading principal minor confirms that the quadratic form remains positive as you add dimensions. If any minor fails (≤ 0), the matrix "collapses" into indefiniteness or negative definiteness.
This criterion is especially efficient when verifying positive definiteness of Hessian matrices in optimization or checking convergence conditions in numerical methods.
Key Takeaway: Sylvester's Criterion lets you verify positive definiteness by checking that all leading principal minors are positive—a purely determinant-based test that avoids eigenvalue computation.