It is named after Jørgen Pedersen Gram.
For finite-dimensional real vectors in with the usual Euclidean dot product, the Gram matrix is simply , where is a matrix whose columns are the vectors . For complex vectors in , , where is the conjugate transpose of .
Given square-integrable functions on the interval , the Gram matrix is:
- In Riemannian geometry, given an embedded -dimensional Riemannian manifold and a coordinate chart for , the volume form on induced by the embedding may be computed using the Gramian of the coordinate tangent vectors:
This generalizes the classical surface integral of a parametrized surface for :
- If the vectors are centered random variables, the Gramian is approximately proportional to the covariance matrix, with the scaling determined by the number of elements in the vector.
- In quantum chemistry, the Gram matrix of a set of basis vectors is the overlap matrix.
- In control theory (or more generally systems theory), the controllability Gramian and observability Gramian determine properties of a linear system.
- Gramian matrices arise in covariance structure model fitting (see e.g., Jamshidian and Bentler, 1993, Applied Psychological Measurement, Volume 18, pp. 79–94).
- In the finite element method, the Gram matrix arises from approximating a function from a finite dimensional space; the Gram matrix entries are then the inner products of the basis functions of the finite dimensional subspace.
- In machine learning, kernel functions are often represented as Gram matrices.
- Since the Gram matrix over the reals is a symmetric matrix, it is diagonalizable and its eigenvalues are non-negative. The diagonalization of the Gram matrix is the singular value decomposition.
The Gramian matrix is positive-semidefinite, and every positive symmetric semidefinite matrix is the Gramian matrix for some set of vectors. Further, in finite-dimensions it determines the vectors up to isomorphism, i.e. any two sets of vectors with the same Gramian matrix must be related by a single unitary matrix. These facts follow from taking the spectral decomposition of any positive-semidefinite matrix , so that and so is the Gramian matrix of the rows of . The Gramian matrix of any orthonormal basis is the identity matrix. The infinite-dimensional analog of this statement is Mercer's theorem.
Derivation of positive-semidefiniteness
The fact that the Gramian matrix is positive-semidefinite can be seen from the following simple derivation:
The first equality follows from the definition of matrix multiplication, the second and third from the bi-linearity of the inner-product, and the last from the positive definiteness of the inner product.
Note that this also shows that the Gramian matrix is positive definite if and only if the vectors are linearly independent.
The Gram determinant or Gramian is the determinant of the Gram matrix:
If are vectors in , then it is the square of the n-dimensional volume of the parallelotope formed by the vectors. In particular, the vectors are linearly independent if and only if the parallelotope has nonzero n-dimensional volume, if and only if Gram determinant is nonzero, if and only if the Gram matrix is nonsingular.
The Gram determinant can also be expressed in terms of the exterior product of vectors by
- Horn & Johnson 2013, p. 441
Theorem 7.2.10 Let be vectors in an inner product space with inner product and let . Then
(a) is Hermitian and positive-semidefinite
(b) is positive-definite if and only if the vectors are linearly-independent.
- Lanckriet, G. R. G.; Cristianini, N.; Bartlett, P.; Ghaoui, L. E.; Jordan, M. I. (2004). "Learning the kernel matrix with semidefinite programming". Journal of Machine Learning Research. 5: 27–72 [p. 29].