Singular value

In mathematics, in particular functional analysis, the singular values, or s-numbers of a compact operator T : X → Y acting between Hilbert spaces X and Y, are the square roots of the eigenvalues of the non-negative self-adjoint operator T*T : X → X (where T* denotes the adjoint of T).

The singular values are non-negative real numbers, usually listed in decreasing order (s₁(T), s₂(T), …). The largest singular value s₁(T) is equal to the operator norm of T (see Min-max theorem).

In the case that T acts on euclidean space Rⁿ, there is a simple geometric interpretation for the singular values: Consider the image by T of the unit sphere; this is an ellipsoid, and the lengths of its semi-axes are the singular values of T (the figure provides an example in R²).

In the case of a normal matrix A, the spectral theorem can be applied to obtain unitary diagonalization of A as per A = UΛU*. Therefore, A ∗ A = U | Λ | U ∗ and so the singular values are simply the absolute values of the eigenvalues.

Most norms on Hilbert space operators studied are defined using s-numbers. For example, the Ky Fan-k-norm is the sum of first k singular values, the trace norm is the sum of all singular values, and the Schatten norm is the pth root of the sum of the pth powers of the singular values. Note that each norm is defined only on a special class of operators, hence s-numbers are useful in classifying different operators.

In the finite-dimensional case, a matrix can always be decomposed in the form UΣV*, where U and V* are unitary matrices and Σ is a diagonal matrix with the singular values lying on the diagonal. This is the singular value decomposition.