Convergence analysis of parametric identification methods for jump processes

Abstract

The asymptotic properties of recursive estimate for the parameter of the signal process of a jump process are investigated. When increasing the observation period, conditions are proposed that guarantee the consistency and asymptotic normality of the estimate. An estimate of the unknown parameter for the random intensity of an extended gamma process model is discussed by way of example.

Keywords:

• Filtering algorithm
• Jump processes
• Extended gamma processes