Ornstein–Uhlenbeck operator

In mathematics, the Ornstein–Uhlenbeck operator is a generalization of the Laplace operator to an infinite-dimensional setting. The Ornstein–Uhlenbeck operator plays a significant role in the Malliavin calculus.

The Laplacian
Consider the gradient operator &nabla; acting on scalar functions f : Rn → R; the gradient of a scalar function is a vector field v = &nabla;f : Rn → Rn. The divergence operator div, acting on vector fields to produce scalar fields, is the adjoint operator to &nabla;. The Laplace operator &Delta; is then the composition of the divergence and gradient operators:


 * $$\Delta = \mathrm{div} \circ \nabla$$,

acting on scalar functions to produce scalar functions. Note that A = &minus;&Delta; is a positive operator, whereas &Delta; is a dissipative operator.

Using spectral theory, one can define a square root (1 &minus; &Delta;)1/2 for the operator (1 &minus; &Delta;). This square root satisfies the following relation involving the Sobolev H1-norm and L2-norm for suitable scalar functions f:


 * $$\big\| f \big\|_{H^{1}}^{2} = \big\| (1 - \Delta)^{1/2} f \big\|_{L^{2}}^{2}.$$

The Ornstein–Uhlenbeck operator
Often, when working on Rn, one works with respect to Lebesgue measure, which has many nice properties. However, remember that the aim is to work in infinite-dimensional spaces, and it is a fact that there is no infinite-dimensional Lebesgue measure. Instead, if one is studying some separable Banach space E, what does make sense is a notion of Gaussian measure; in particular, the abstract Wiener space construction makes sense.

To get some intuition about what can be expected in the infinite-dimensional setting, consider standard Gaussian measure &gamma;n on Rn: for Borel subsets A of Rn,


 * $$\gamma^{n} (A) := \int_{A} (2 \pi)^{-n/2} \exp ( - | x |^{2} / 2) \, \mathrm{d} x.$$

This makes (Rn, B(Rn), &gamma;n) into a probability space; E will denote expectation with respect to &gamma;n.

The gradient operator &nabla; acts on a (differentiable) function &phi; : Rn → R to give a vector field &nabla;&phi; : Rn → Rn.

The divergence operator &delta; (to be more precise, &delta;n, since it depends on the dimension) is now defined to be the adjoint of &nabla; in the Hilbert space sense, in the Hilbert space L2(Rn, B(Rn), &gamma;n; R). In other words, &delta; acts on a vector field v : Rn → Rn to give a scalar function &delta;v : Rn → R, and satisfies the formula


 * $$\mathbb{E} \big[ \nabla f \cdot v \big] = \mathbb{E} \big[ f \delta v \big].$$

On the left, the product is the pointwise Euclidean dot product of two vector fields; on the right, it is just the pointwise multiplication of two functions. Using integration by parts, one can check that &delta; acts on a vector field v with components vi, i = 1, ..., n, as follows:


 * $$\delta v (x) = \sum_{i = 1}^{n} \left( x_{i} v^{i} (x) - \frac{\partial v^{i}}{\partial x_{i}} (x) \right).$$

The change of notation from "div" to "&delta;" is for two reasons: first, &delta; is the notation used in infinite dimensions (the Malliavin calculus);  secondly, &delta; is really the negative of the usual divergence.

The (finite-dimensional) Ornstein–Uhlenbeck operator L (or, to be more precise, Lm) is defined by


 * $$L := - \delta \circ \nabla,$$

with the useful formula that for any f and g smooth enough for all the terms to make sense,


 * $$\delta ( f \nabla g) = - \nabla f \cdot \nabla g - f L g.$$

The Ornstein–Uhlenbeck operator L is related to the usual Laplacian &Delta; by


 * $$L f (x) = \Delta f (x) - x \cdot \nabla f (x).$$

The Ornstein–Uhlenbeck operator for a separable Banach space
Consider now an abstract Wiener space E with Cameron-Martin Hilbert space H and Wiener measure &gamma;. Let D denote the Malliavin derivative. The Malliavin derivative D is an unbounded operator from L2(E, &gamma;; R) into L2(E, &gamma;; H) – in some sense, it measures "how random" a function on E is. The domain of D is not the whole of L2(E, &gamma;; R), but is a dense linear subspace, the Watanabe-Sobolev space, often denoted by $$\mathbb{D}^{1,2}$$ (once differentiable in the sense of Malliavin, with derivative in L2).

Again, &delta; is defined to be the adjoint of the gradient operator (in this case, the Malliavin derivative is playing the role of the gradient operator). The operator &delta; is also known the Skorokhod integral, which is an anticipating stochastic integral; it is this set-up that gives rise to the slogan "stochastic integrals are divergences". &delta; satisfies the identity


 * $$\mathbb{E} \big[ \langle \mathrm{D} F, v \rangle_{H} \big] = \mathbb{E} \big[ F \delta v \big]$$

for all F in $$\mathbb{D}^{1,2}$$ and v in the domain of &delta;.

Then the Ornstein–Uhlenbeck operator for E is the operator L defined by


 * $$L := - \delta \circ \mathrm{D}.$$