Scandinavian Journal of Statistics · 2000 · 84 citations · 9 references
EngineeringMathematicsStochastic ModelingStochastic ProcessesTime Series AnalysisBrownian MotionDiffusion ProcessesDiscrete-time Stochastic ProcessesErgodic TheoryStochastic CalculusStochastic Differential EquationsDiffusion Coefficient EstimationErgodic PropertiesErgodicity Of DiffusionsInference For Diffusion ProcessesMinimum Variance EstimationParameter Estimation For DiffusionsParameter Estimation For Stochastic Differential EquationsSimulation Of Diffusion Processes
We consider a one‐dimensional diffusion process X , with ergodic property, with drift b ( x , θ) and diffusion coefficient a ( x , θ) depending on an unknown parameter θ that may be multidimensional. We are interested in the estimation of θ and dispose, for that purpose, of a discretized trajectory, observed at n equidistant times t i = iΔ , i = 0, ..., n . We study a particular class of estimating functions of the form ∑ f (θ, X t i −1 ) which, under the assumption that the integral of f with respect to the invariant measure is null, provide us with a consistent and asymptotically normal estimator. We determine the choice of f that yields the estimator with minimum asymptotic variance within the class and indicate how to construct explicit estimating functions based on the generator of the diffusion. Finally the theoretical study is completed with simulations.
9