Expected A Posteriori Estimation in FinancialApplications

This paper introduces a new method for online estimation and predictionof states and parameters of nonlinear stochastic differential equations. Inthis setup parameters are considered as random variables in a Bayesiansense, which requires integration over parameter distributions. This isaccomplished by well suited quadratures.The suggested procedure is incorporated into a state space architecture,which allows for sequential calculation of likelihood functions. Thisis done by normal correlation updates and prediction error decomposition.The resulting EAP-Filter can process a variety of nonlinear problems, includinglatent states. Additionally, estimates and predictions for systemstates and parameters can be calculated online, without iterative loops.....