On stochastic differential equations and semigroups of probability operators in quantum probability (Q1965902): Difference between revisions

A Hilbert-space-valued stochastic differential equation of the form \[ d\psi _t=-K_t\psi _{t-}dt+\sum _{k=1}^\infty L_{kt}\psi _{t-}dW_{kt}+\int _{\mathcal Y} (J_t\psi _{t-})(y)\tilde \Pi (dy,dt) \] with an initial condition \(\psi _0=\xi \) is studied where \(\xi \in \mathcal H\), \(L_t\), \(J_t\) and \(K_t\) are measurable functions with values in \(\mathcal L(\mathcal H;\mathcal H \otimes l_2), \mathcal L(\mathcal H; L^2(\mathcal Y,\nu ,\mathcal H))\) and \( \mathcal L(\mathcal H)\), respectively, \(\mathcal H\) denotes a separable complex Hilbert space, \(\mathcal Y\) is a locally compact Hausdorff space with a topology with a countable basis, and \(\nu \) is a Radon measure on its Borel \(\sigma \)-field. \(W_{kt}\) denotes a sequence of adapted, standard, independent Wiener processes and \(\Pi (dy,dt)\) is an adapted Poisson point process on \(\mathcal Y\times R_+\).The above equation is investigated in connection with the approach to continuous measurements based on semigroups of operators arising in quantum probability. A stochastic representation of a general semigroup of such probability operators is obtained. The method allows to treat also non-homogeneous cases when the semigroups are replaced by evolution systems.