{
  "id": "sylvesters-criterion",
  "title": "Sylvester's criterion",
  "lang": "en",
  "source_lang": "en",
  "summary": "Sylvester's criterion characterizes positive definiteness of a real symmetric or complex Hermitian matrix by positivity of its leading principal minors.",
  "category": "mathematics",
  "subcategory": "algebra",
  "type": "concept",
  "aliases": [
    "Sylvester criterion",
    "Sylvester's positive-definiteness criterion"
  ],
  "generated_at": "2026-11-05",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/sylvesters-criterion/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/sylvesters-criterion/",
      "json": "https://aiwiki.page/zh/sylvesters-criterion.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "Sylvester's criterion is a theorem in linear algebra stating that a real symmetric or complex Hermitian matrix is positive definite if and only if all its leading principal minors are positive. It converts a condition on a quadratic form evaluated at every nonzero vector into a finite list of determinant inequalities. The criterion is named after James Joseph Sylvester. Related tests characterize negative definiteness and positive semidefiniteness, but the semidefinite case requires checking all principal minors, not merely the leading ones."
    },
    {
      "heading": "Definitions and statement",
      "text": "Let \\(A\\) be an \\(n\\times n\\) Hermitian matrix, meaning \\(A=A^*\\), where \\(A^*\\) denotes its conjugate transpose. For real matrices, this condition reduces to symmetry, \\(A=A^{\\mathsf T}\\).\n\nThe associated quadratic form is\n\n\\[q(x)=x^*Ax.\\]\n\nBecause \\(A\\) is Hermitian, \\(q(x)\\) is real. Positive definiteness means\n\n\\[x^*Ax>0 \\qquad\\text{for every }x\\ne 0.\\]\n\nA principal submatrix is obtained by selecting the same set of row and column indices. Its determinant is a principal minor. The leading principal submatrix of order \\(k\\) uses the first \\(k\\) rows and columns:\n\n\\[A_k=A[\\{1,\\ldots,k\\},\\{1,\\ldots,k\\}].\\]\n\nWrite\n\n\\[\\Delta_k=\\det A_k,\\qquad \\Delta_0=1.\\]\n\nSylvester's criterion states:\n\n\\[\\boxed{ A\\text{ is positive definite} \\iff \\Delta_k>0\\quad\\text{for }k=1,\\ldots,n. }\\]\n\nOnly \\(n\\) determinants appear in this test. By contrast, there are \\(2^n-1\\) nonempty principal index sets.\n\nThe symmetry or Hermitian assumption is essential. For an arbitrary real matrix,\n\n\\[x^{\\mathsf T}Ax = x^{\\mathsf T}\\left(\\frac{A+A^{\\mathsf T}}{2}\\right)x,\\]\n\nso positivity of its quadratic form is determined by its symmetric part, not generally by the leading minors of \\(A\\) itself."
    },
    {
      "heading": "Examples",
      "text": "A two-dimensional matrix\n\nFor a real symmetric matrix\n\n\\[A= \\begin{pmatrix} a&b\\\\ b&c \\end{pmatrix},\\]\n\nthe criterion gives\n\n\\[A\\text{ is positive definite} \\iff a>0 \\quad\\text{and}\\quad ac-b^2>0.\\]\n\nThese conditions also imply \\(c>0\\). Positive diagonal entries alone are not sufficient: the matrix\n\n\\[\\begin{pmatrix} 1&2\\\\ 2&1 \\end{pmatrix}\\]\n\nhas positive diagonal entries but determinant \\(-3\\), and its quadratic form takes both positive and negative values.\n\nFor the complex Hermitian matrix\n\n\\[A= \\begin{pmatrix} a&z\\\\ \\overline z&c \\end{pmatrix}, \\qquad a,c\\in\\mathbb R,\\]\n\nthe corresponding conditions are\n\n\\[a>0,\\qquad ac-|z|^2>0.\\]\n\nA three-dimensional matrix\n\nConsider\n\n\\[A= \\begin{pmatrix} 2&-1&0\\\\ -1&2&-1\\\\ 0&-1&2 \\end{pmatrix}.\\]\n\nIts leading principal minors are\n\n\\[\\Delta_1=2,\\qquad \\Delta_2=3,\\qquad \\Delta_3=4.\\]\n\nAll are positive, so \\(A\\) is positive definite without explicitly computing its eigenvalues."
    },
    {
      "heading": "Why the criterion works",
      "text": "Necessity\n\nSuppose \\(A\\) is positive definite. Every leading principal submatrix \\(A_k\\) is also positive definite: for any nonzero \\(y\\in\\mathbb C^k\\), extend \\(y\\) to a vector in \\(\\mathbb C^n\\) by appending zeros. Then\n\n\\[y^*A_ky>0.\\]\n\nBy the spectral theorem, a Hermitian positive-definite matrix has strictly positive eigenvalues. Its determinant is their product, so\n\n\\[\\Delta_k>0\\]\n\nfor every \\(k\\).\n\nSufficiency\n\nThe converse follows by induction and completing the square. The one-dimensional case is immediate. For the induction step, write\n\n\\[A= \\begin{pmatrix} B&b\\\\ b^*&c \\end{pmatrix},\\]\n\nwhere \\(B=A_{n-1}\\). The leading principal minors of \\(B\\) are positive, so the induction hypothesis makes \\(B\\) positive definite and invertible.\n\nDefine the scalar\n\n\\[s=c-b^*B^{-1}b.\\]\n\nThe block determinant identity gives\n\n\\[\\det A=(\\det B)s,\\]\n\nand therefore\n\n\\[s=\\frac{\\Delta_n}{\\Delta_{n-1}}>0.\\]\n\nFor \\(u\\in\\mathbb C^{n-1}\\) and \\(t\\in\\mathbb C\\),\n\n\\[\\begin{pmatrix}u\\\\t\\end{pmatrix}^* A \\begin{pmatrix}u\\\\t\\end{pmatrix} = (u+B^{-1}bt)^*B(u+B^{-1}bt)+s|t|^2.\\]\n\nBoth terms are nonnegative. If \\(t\\ne0\\), the second is positive; if \\(t=0\\) and \\(u\\ne0\\), the first is positive. Thus \\(A\\) is positive definite."
    },
    {
      "heading": "Factorization interpretation",
      "text": "When every leading principal minor is nonzero, a Hermitian matrix admits a factorization\n\n\\[A=LDL^*,\\]\n\nwhere \\(L\\) is unit lower triangular and \\(D\\) is real diagonal. Its diagonal entries satisfy\n\n\\[d_k=\\frac{\\Delta_k}{\\Delta_{k-1}}.\\]\n\nConsequently, positivity of every \\(\\Delta_k\\) is equivalent to positivity of every diagonal entry of \\(D\\). Since\n\n\\[x^*Ax=(L^*x)^*D(L^*x),\\]\n\nand \\(L^*\\) is invertible, this gives another proof of the criterion.\n\nWhen \\(D\\) is positive, its diagonal entries have positive square roots, yielding\n\n\\[A=CC^*, \\qquad C=LD^{1/2}.\\]\n\nThis is the Cholesky decomposition. Sylvester's criterion therefore connects determinant inequalities with the existence of a triangular factorization having positive diagonal entries."
    },
    {
      "heading": "Negative definiteness and semidefiniteness",
      "text": "Negative definiteness\n\nA Hermitian matrix is negative definite when\n\n\\[x^*Ax<0 \\qquad(x\\ne0).\\]\n\nApplying Sylvester's criterion to \\(-A\\) gives\n\n\\[A\\text{ is negative definite} \\iff (-1)^k\\Delta_k>0 \\quad(k=1,\\ldots,n).\\]\n\nThus its leading principal minors alternate in sign:\n\n\\[\\Delta_1<0,\\quad \\Delta_2>0,\\quad \\Delta_3<0,\\quad\\ldots\\]\n\nPositive semidefiniteness\n\nA Hermitian matrix is positive semidefinite when\n\n\\[x^*Ax\\ge0 \\qquad\\text{for every }x.\\]\n\nThe correct principal-minor characterization is\n\n\\[A\\text{ is positive semidefinite} \\iff \\text{every principal minor of }A\\text{ is nonnegative}.\\]\n\nSimply replacing \\(>\\) by \\(\\ge\\) in the leading-minor test is invalid. For example,\n\n\\[A= \\begin{pmatrix} 0&0\\\\ 0&-1 \\end{pmatrix}\\]\n\nhas leading principal minors \\(0\\) and \\(0\\), yet\n\n\\[\\begin{pmatrix}0&1\\end{pmatrix} A \\begin{pmatrix}0\\\\1\\end{pmatrix} =-1.\\]\n\nIts nonleading principal minor \\(a_{22}=-1\\) detects the failure.\n\nNecessity follows because every principal submatrix of a positive-semidefinite matrix is positive semidefinite. For sufficiency, if all principal minors are nonnegative, then every leading principal minor of\n\n\\[A+\\varepsilon I,\\qquad \\varepsilon>0,\\]\n\nis strictly positive: expanding its determinant in powers of \\(\\varepsilon\\) gives nonnegative coefficients from principal minors and a positive highest-degree term. Sylvester's criterion makes \\(A+\\varepsilon I\\) positive definite. Letting \\(\\varepsilon\\) tend to zero yields \\(x^*Ax\\ge0\\).\n\nSimilarly, negative semidefiniteness requires every principal minor of order \\(k\\) to satisfy\n\n\\[(-1)^k\\det A[I,I]\\ge0.\\]"
    },
    {
      "heading": "Applications and computational limitations",
      "text": "In mathematical optimization, the criterion supplies a practical test for the Hessian matrix at a stationary point. For a twice continuously differentiable real-valued function, a positive-definite Hessian there implies a strict local minimum; a negative-definite Hessian implies a strict local maximum. A semidefinite Hessian alone generally does not determine the type of stationary point.\n\nIn statistics, covariance matrices are positive semidefinite and are positive definite when no nonzero linear combination of the variables has zero variance. The criterion can establish strict positive definiteness in symbolic calculations. Related uses arise with Gram matrices, quadratic objectives, and finite-element stiffness matrices.\n\nFor small matrices or matrices depending on symbolic parameters, determinant inequalities are often convenient. For large numerical problems, explicitly evaluating every leading determinant is usually less suitable than attempting Cholesky factorization. A standard dense Cholesky factorization takes \\(O(n^3)\\) arithmetic operations and avoids separately recomputing determinants.\n\nIn floating-point arithmetic, rounding can obscure the signs of very small minors. A nearly singular positive-definite matrix may therefore be difficult to distinguish numerically from a semidefinite or slightly indefinite matrix. The exact criterion remains valid, but numerical conclusions depend on scaling, conditioning, and error tolerances.\n\nFinally, failure of a leading-minor test proves only that the matrix is not positive definite. It does not by itself show whether the matrix is indefinite, negative definite, or singular."
    }
  ],
  "links": [
    "linear-algebra",
    "matrix",
    "positive-definite-matrix",
    "hermitian-matrix",
    "conjugate-transpose",
    "quadratic-form",
    "determinant",
    "eigenvalues-and-eigenvectors",
    "spectral-theorem",
    "mathematical-induction",
    "cholesky-decomposition",
    "mathematical-optimization",
    "hessian-matrix",
    "covariance-matrix",
    "gram-matrix",
    "finite-element-method",
    "floating-point-arithmetic",
    "condition-number"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}