A Note on Interpolation of Stable Processes
AbstractInterpolation procedures tailored for gaussian processes may not be applied to infinite variance stable processes. Alternative techniques suitable for a limited set of stable case with index α∈(1,2] were initially studied by Pourahmadi (1984) for harmonizable processes. This was later extended to the ARMA stable process with index α∈(0,2] by Nassiuma (1994). In this paper, the problem of interpolation of stable processes is studied with the aim of developing an algorithm applicable to general linear and nonlinear processes by using the state space formulation. Application of this procedure to the estimation of missing values is discussed.
Journal of Agriculture, Science and Technology Vol.3(1) 2001: 81-87