1 Statement of the criterion

Sylvester’s criterion gives a determinant-based test for positive definiteness of a real symmetric matrix. It states that a symmetric matrix is positive definite if and only if all of its leading principal minors are positive. This result is widely used because it replaces a global matrix condition with a finite sequence of smaller calculations.

1.1 Positive definite matrices

A real symmetric matrix is positive definite when its associated quadratic form is strictly positive for every nonzero vector. In practical terms, this means the matrix bends the geometry of space in a uniformly inward way, never producing zero or negative values except at the origin. Such matrices arise naturally in optimization, geometry, and numerical analysis.

1.2 Leading principal minors

A leading principal minor is the determinant of the upper-left k × k submatrix of a matrix, for each k from 1 up to the full size. These minors are called “leading” because they are taken from the top-left corner in order. Sylvester’s criterion requires checking each of these determinants consecutively.

1.3 Necessary and sufficient conditions

For a real symmetric matrix, the positivity of all leading principal minors is not merely helpful but exact: it is both necessary and sufficient for positive definiteness. If even one leading principal minor is zero or negative, the matrix fails the criterion. Conversely, if every one is positive, the matrix is positive definite.

1.4 Relation to symmetric matrices

The criterion applies in its standard form to symmetric matrices because symmetry guarantees real quadratic forms and a close connection with eigenvalues. For nonsymmetric matrices, leading principal minors do not provide the same characterization. Symmetry is therefore an essential hypothesis in the classical statement.

2 Background and motivation

Sylvester’s criterion is rooted in the study of quadratic forms, where one seeks to classify expressions built from sums of products of variables. The need for efficient tests of definiteness became especially important as matrix methods spread through algebra, mechanics, and optimization. The criterion gives an algebraic route to a geometric question.

2.1 Quadratic forms

A quadratic form is an expression of the form x^T A x, where A is a matrix and x is a vector. When A is symmetric, the form is determined by the entries of A in a structured way. Determining whether such a form is always positive is a central problem in classical algebra.

2.2 Definiteness and eigenvalues

Positive definiteness can also be described through eigenvalues: a symmetric matrix is positive definite exactly when all eigenvalues are positive. Sylvester’s criterion offers an alternative that avoids computing eigenvalues directly. This can be advantageous when determinants of small submatrices are easier to evaluate than spectral data.

2.3 Historical development

The criterion is named after James Joseph Sylvester, who made major contributions to matrix theory and invariant methods in algebra. Its development reflects nineteenth-century work on forms and determinants. Over time, it became a standard tool in linear algebra textbooks and applications.

3 Proofs and derivations

Several standard proofs establish Sylvester’s criterion from different perspectives. Each proof highlights a different structural feature of positive definite matrices. Together, they show why the criterion is both natural and robust.

3.1 Proof via induction

An inductive proof examines the matrix one leading block at a time. The first minor guarantees the first diagonal entry is positive, and successive minors control the remaining entries after suitable elimination. At each step, positivity of the next minor ensures the reduced smaller matrix remains positive definite.

3.2 Proof using Gaussian elimination

Gaussian elimination can transform a symmetric positive definite matrix into an upper triangular or diagonal form without changing the relevant sign structure. The pivots that appear during elimination are ratios of successive leading principal minors. Positivity of all minors therefore forces all pivots to be positive, which implies positive definiteness.

3.3 Proof using Cholesky decomposition

A symmetric matrix is positive definite exactly when it admits a Cholesky factorization A = LL^T with L lower triangular and positive diagonal entries. The diagonal entries of L are determined by successive leading principal minors. Positivity of these minors guarantees that the factorization exists, which in turn proves positive definiteness.

3.4 Proof via principal submatrices

One can also argue using the behavior of principal submatrices. If all leading principal minors are positive, then each leading principal submatrix is itself positive definite. This nested structure implies that the full matrix cannot fail positive definiteness without violating positivity at some earlier stage.

4 Applications

Sylvester’s criterion is most useful when one needs a quick algebraic test rather than a full eigenvalue computation. It appears in settings where matrix positivity controls existence, uniqueness, or stability. Its role is especially prominent in theoretical and computational work with symmetric matrices.

4.1 Testing positive definiteness

In many problems, the first question is whether a symmetric matrix is positive definite. Sylvester’s criterion provides a direct check using determinants of successive upper-left blocks. This is especially effective for small and medium-sized matrices.

4.2 Matrix optimization

In optimization, positive definite Hessian matrices indicate local minima of smooth functions. Sylvester’s criterion can verify this condition by checking minors of the Hessian. As a result, it is often used in multivariable calculus and constrained optimization.

4.3 Stability analysis

In dynamical systems and mechanics, positive definite matrices often represent energy or stiffness. Their definiteness helps determine whether an equilibrium is stable or whether an energy function has a local minimum. Sylvester’s criterion supplies a convenient test in these analyses.

4.4 Quadratic form classification

Quadratic forms can be classified by whether they are positive definite, negative definite, or indefinite. Sylvester’s criterion helps identify the positive definite case directly from coefficients. Related sign patterns can be used to classify other types of forms.

Sylvester’s criterion belongs to a broader collection of results about matrix signatures, minors, and factorization. These results are closely connected and often used together. They provide multiple ways to analyze the same underlying structure.

5.1 Sylvester’s law of inertia

Sylvester’s law of inertia states that a real symmetric matrix has an invariant count of positive, negative, and zero eigenvalues under congruence transformations. This law complements Sylvester’s criterion by describing the full signature of a quadratic form. Together, they form a central part of matrix classification theory.

5.2 Principal minor tests

Other determinant-based tests use principal minors rather than only leading ones. These variants can address semidefinite or more general sign conditions, though they are usually less simple than Sylvester’s criterion. The leading-minor test remains the most streamlined for positive definiteness.

5.3 Criteria for positive semidefinite matrices

Positive semidefinite matrices allow zero values in their quadratic forms, so the strict inequalities of Sylvester’s criterion no longer suffice. Testing semidefiniteness typically requires checking all principal minors or using eigenvalue methods. This makes the semidefinite case more delicate.

5.4 Cholesky factorization

Cholesky factorization is a decomposition available precisely for positive definite matrices. It is closely tied to the criterion because the existence of a Cholesky factor reflects the same positivity conditions. In computation, it is often used both as a test and as a numerical tool.

6 Examples

Examples make the criterion straightforward to apply and show how quickly it can detect definiteness. The calculations are simple in small dimensions, but they already illustrate the full pattern. Careful attention to the order of the leading minors is essential.

6.1 2×2 matrices

For a symmetric matrix [[a, b], [b, c]], Sylvester’s criterion requires a > 0 and ac - b^2 > 0. The first condition checks the 1 × 1 leading minor, and the second checks the full determinant. If both hold, the matrix is positive definite.

6.2 3×3 matrices

For a 3 × 3 symmetric matrix, one checks three leading minors in order. First, the top-left entry must be positive; second, the determinant of the upper-left 2 × 2 block must be positive; third, the determinant of the full matrix must be positive. All three conditions together guarantee positive definiteness.

6.3 Computing leading principal minors

To compute the minors, one extracts each upper-left block and evaluates its determinant. In practice, symbolic simplification or row reduction may help. The minors often reveal a clear pattern, especially for matrices with repeated or structured entries.

6.4 Common pitfalls and counterexamples

A common mistake is to check all diagonal entries instead of the leading principal minors. Positive diagonal entries alone do not imply positive definiteness. Another error is to confuse principal minors with arbitrary subdeterminants, which are not sufficient in the same direct way.

7 Extensions and variants

Sylvester’s criterion has several useful extensions, though each comes with its own limits. Some variants change the sign pattern, while others adapt the idea to complex matrices or block-structured systems. These generalizations preserve the same underlying philosophy.

7.1 Negative definite matrices

A symmetric matrix is negative definite when its quadratic form is always negative for nonzero vectors. This can be tested by applying Sylvester-type sign conditions to the leading principal minors, with alternating signs rather than all positive ones. The pattern reflects the effect of multiplication by -1.

7.2 Positive semidefinite limitations

For semidefinite matrices, the criterion no longer works with simple strict inequalities. Zero minors may occur even when the matrix is semidefinite, so additional information is required. As a result, semidefinite testing is usually broader and less elementary.

7.3 Hermitian matrix analogues

In complex linear algebra, symmetric matrices are replaced by Hermitian matrices. A parallel criterion holds for Hermitian positive definite matrices, using leading principal minors computed in the complex setting. The same conceptual link between minors and definiteness remains intact.

7.4 Block matrix versions

Block matrix analogues extend the idea by replacing scalar minors with determinants of leading block submatrices. These versions are useful when a matrix has natural partitioning into subspaces or coupled systems. They are common in numerical methods and structured matrix analysis.