Three kinds of estimates for performance sensitivities (gradients, Hessians etc.) of stochastic systems are introduced. These estimates are given in general operator form. Their convergence conditions and rate of convergence are presented. Particular attention is given to estimates obtained from a single sample path. Various examples of estimates are considered. © 1988.
Kreimer, J. (1988). Generalized estimates for performance sensitivities of stochastic systems. Mathematical and Computer Modelling, 10(12), 911–922. https://doi.org/10.1016/0895-7177(88)90183-5